{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "45111406-ed47-49b8-ade4-22877cb3313d",
   "metadata": {},
   "source": [
    "# NMF Topic Modeling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f31c926f-8e9c-4574-b440-0894488142c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.datasets import fetch_20newsgroups\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.decomposition import NMF\n",
    "from nltk.corpus import stopwords # nltk = package for human language data processing. \n",
    "import nltk\n",
    "import os\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "0b15f3cd-69a8-4dcb-8684-27fbd5985144",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "New working directory: C:\\Users\\vgonz\\OneDrive - University of Pittsburgh\\Dissertation - Vale\\Paper 2 - Political-Economic Polarization\\Publication\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# Get the current working directory (where the Jupyter notebook is located)\n",
    "current_directory = os.getcwd()\n",
    "\n",
    "# Move one level up\n",
    "parent_directory = os.path.dirname(current_directory)\n",
    "\n",
    "# Change the working directory to the parent directory\n",
    "os.chdir(parent_directory)\n",
    "print(\"New working directory:\", os.getcwd())\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b71ef513-68f9-462a-8875-a935ddd8038e",
   "metadata": {
    "tags": []
   },
   "source": [
    "# All the text\n",
    "\n",
    "### Loading data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "43473830-18ea-45dc-8843-85fd81f2a1e1",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C:\\Users\\vgonz\\OneDrive - University of Pittsburgh\\Dissertation - Vale\\Paper 2 - Political-Economic Polarization\\Publication\n"
     ]
    }
   ],
   "source": [
    "# Importing modules\n",
    "import pandas as pd\n",
    "import os\n",
    "print(os.getcwd())\n",
    "# Read data into papers\n",
    "papers = pd.read_csv(\"Data/cleaned_data_G.csv\")\n",
    "# Print head\n",
    "papers.head()\n",
    "papers['text_new'] = papers['clean_text']\n",
    "\n",
    "papers = papers[['text_new']]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "5b2730ef-6403-4e4b-9245-46602ab4bc80",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# regular expression to remove any punctuation, and then lowercase the text\n",
    "# Load the regular expression library\n",
    "import re\n",
    "# Remove punctuation\n",
    "papers['text_new'] = \\\n",
    "papers['text_new'].map(lambda x: str(x))\n",
    "papers['text_new'] = \\\n",
    "papers['text_new'].map(lambda x: re.sub('[,\\.!?]', '', x))\n",
    "# Convert the titles to lowercase\n",
    "papers['text_new'] = \\\n",
    "papers['text_new'].map(lambda x: x.lower())\n",
    "\n",
    "# Print out the first rows of papers\n",
    "papers['text_new'].head()\n",
    "\n",
    "data = papers.text_new.values.tolist()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "324b2970-cbdc-4011-bae2-5c119750c660",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  (0, 93)\t0.07174913158727336\n",
      "  (0, 141)\t0.08517716772192677\n",
      "  (0, 192)\t0.08062221260594189\n",
      "  (0, 28)\t0.14992129008730995\n",
      "  (0, 112)\t0.07330763289579756\n",
      "  (0, 85)\t0.0786015587106631\n",
      "  (0, 75)\t0.11044404142329659\n",
      "  (0, 130)\t0.12344523816910023\n",
      "  (0, 90)\t0.0786015587106631\n",
      "  (0, 37)\t0.08280468507616007\n",
      "  (0, 21)\t0.07496064504365497\n",
      "  (0, 59)\t0.08280468507616007\n",
      "  (0, 179)\t0.06506553689091266\n",
      "  (0, 169)\t0.07496064504365497\n",
      "  (0, 124)\t0.08280468507616007\n",
      "  (0, 106)\t0.16560937015232013\n",
      "  (0, 34)\t0.1572031174213262\n",
      "  (0, 150)\t0.08517716772192677\n",
      "  (0, 74)\t0.0786015587106631\n",
      "  (0, 166)\t0.08280468507616007\n",
      "  (0, 175)\t0.06887633490716778\n",
      "  (0, 44)\t0.14992129008730995\n",
      "  (0, 35)\t0.08280468507616007\n",
      "  (0, 116)\t0.05522202071164829\n",
      "  (0, 23)\t0.08517716772192677\n",
      "  :\t:\n",
      "  (88, 72)\t0.08947727669259041\n",
      "  (88, 60)\t0.08624589213964215\n",
      "  (88, 108)\t0.2645719464032926\n",
      "  (88, 170)\t0.06802572502537824\n",
      "  (88, 102)\t0.07387435634840024\n",
      "  (88, 47)\t0.06595541571635179\n",
      "  (88, 73)\t0.048581700363100805\n",
      "  (89, 83)\t0.24648434231513133\n",
      "  (89, 153)\t0.22745594884055537\n",
      "  (89, 41)\t0.20762650348631298\n",
      "  (89, 177)\t0.24648434231513133\n",
      "  (89, 80)\t0.23961888951556226\n",
      "  (89, 11)\t0.23330328516019314\n",
      "  (89, 123)\t0.19546356281130609\n",
      "  (89, 121)\t0.19179299716085516\n",
      "  (89, 125)\t0.15980061113160587\n",
      "  (89, 152)\t0.4242729399030788\n",
      "  (89, 45)\t0.5451197078575462\n",
      "  (89, 108)\t0.21691993039053534\n",
      "  (89, 154)\t0.21691993039053534\n",
      "  (90, 2)\t0.3314975196010503\n",
      "  (90, 123)\t0.2918564098205376\n",
      "  (90, 7)\t0.16430966336553682\n",
      "  (90, 62)\t0.8637687613051982\n",
      "  (90, 73)\t0.17842391113022094\n",
      "X =  ['able' 'according' 'account' 'act' 'addition' 'advocate' 'advocates'\n",
      " 'afd' 'aim' 'allow' 'appropriate' 'areas' 'authorities' 'based' 'basic'\n",
      " 'basis' 'better' 'care' 'case' 'central' 'chapter' 'children' 'citizens'\n",
      " 'civil' 'committed' 'common' 'comprehensive' 'conditions' 'constitution'\n",
      " 'constitutional' 'control' 'core' 'costs' 'countries' 'country' 'create'\n",
      " 'cultural' 'culture' 'current' 'currently' 'decision' 'demand' 'demands'\n",
      " 'democracy' 'democratic' 'development' 'does' 'economic' 'economy' 'end'\n",
      " 'ensure' 'environment' 'equal' 'established' 'eu' 'euro' 'europe'\n",
      " 'european' 'existing' 'families' 'family' 'favour' 'federal' 'financial'\n",
      " 'foreign' 'form' 'framework' 'free' 'freedom' 'fundamental' 'future'\n",
      " 'general' 'german' 'germany' 'good' 'government' 'high' 'immigration'\n",
      " 'income' 'increased' 'increasing' 'individual' 'influence'\n",
      " 'infrastructure' 'instead' 'institutions' 'integration' 'interests'\n",
      " 'international' 'labour' 'language' 'large' 'law' 'laws' 'lead' 'led'\n",
      " 'legal' 'legislation' 'level' 'life' 'limited' 'local' 'long' 'longer'\n",
      " 'low' 'lower' 'maintain' 'make' 'market' 'mass' 'means' 'measures'\n",
      " 'members' 'ment' 'migration' 'nation' 'national' 'nature' 'necessary'\n",
      " 'need' 'needs' 'new' 'non' 'number' 'open' 'order' 'parents' 'parliament'\n",
      " 'particular' 'parties' 'people' 'policies' 'policy' 'political'\n",
      " 'population' 'possible' 'power' 'prevent' 'principle' 'principles'\n",
      " 'private' 'protect' 'protection' 'public' 'quality' 'rates' 'reason'\n",
      " 'reasons' 'reduce' 'reduced' 'reform' 'regard' 'regulations' 'rejects'\n",
      " 'remain' 'required' 'requirements' 'respect' 'result' 'return' 'right'\n",
      " 'rights' 'role' 'schools' 'sector' 'security' 'self' 'services' 'set'\n",
      " 'shall' 'social' 'society' 'special' 'standards' 'state' 'states'\n",
      " 'strengthen' 'sufficient' 'supply' 'support' 'systems' 'taken' 'tax'\n",
      " 'term' 'time' 'tion' 'use' 'used' 'values' 'want' 'wants' 'way' 'welfare'\n",
      " 'whilst' 'years']\n"
     ]
    }
   ],
   "source": [
    "vectorizer = TfidfVectorizer(max_features=1500, min_df=10, stop_words='english')\n",
    "X = vectorizer.fit_transform(data)\n",
    "words = np.array(vectorizer.get_feature_names_out())\n",
    "\n",
    "print(X)\n",
    "print(\"X = \", words)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cdd5f414-dc52-42cf-b572-52bc41a3746f",
   "metadata": {},
   "source": [
    "### 4 topics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "f882d233-c6b5-45b2-accd-91e5d58d9de0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic 1: power,foreign,national,euro,policy,public,political,germany,afd,german\n",
      "Topic 2: foreign,culture,economy,nature,schools,infrastructure,protection,security,policy,chapter\n",
      "Topic 3: special,afd,social,support,schools,care,family,parents,families,children\n",
      "Topic 4: right,legal,germany,country,migration,social,eu,countries,integration,immigration\n"
     ]
    }
   ],
   "source": [
    "nmf = NMF(n_components=4, solver='mu', init='nndsvda')\n",
    "W = nmf.fit_transform(X)\n",
    "H = nmf.components_\n",
    "\n",
    "# Use W to get the most likely topic for each sentence\n",
    "most_likely_topic = W.argmax(axis=1)\n",
    "\n",
    "# Add most_likely_topic as a new column to the papers dataframe\n",
    "papers['most_likely_topic'] = most_likely_topic\n",
    "\n",
    "\n",
    "for i, topic in enumerate(H):\n",
    "     print(\"Topic {}: {}\".format(i + 1, \",\".join([str(x) for x in words[topic.argsort()[-10:]]])))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d0b6d1fb-fb85-4d14-b02c-7f9c71f826b6",
   "metadata": {},
   "source": [
    "### Topics proportions\n",
    "This method calculates the proportion of each topic based on the sum of the weights (or contributions) for each topic across all documents. It gives a sense of how much each topic contributes to the overall data set based on the NMF model's output."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "164f9c10-754c-44bc-9209-1c852c0b668f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[t-SNE] Computing 90 nearest neighbors...\n",
      "[t-SNE] Indexed 91 samples in 0.000s...\n",
      "[t-SNE] Computed neighbors for 91 samples in 0.200s...\n",
      "[t-SNE] Computed conditional probabilities for sample 91 / 91\n",
      "[t-SNE] Mean sigma: 0.087384\n",
      "[t-SNE] KL divergence after 250 iterations with early exaggeration: 46.467045\n",
      "[t-SNE] KL divergence after 300 iterations: 0.075785\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAosAAAKQCAYAAAD3+zBPAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAACTf0lEQVR4nOzdd3yU9eEH8M9zeyWXvRMIYe+NyJKlguKuinWg4kLrrL/WWq2jrbXaqrVq6x4VQVRAFBURVDbIMBDCCmTvncvtu+f3RyQaksu8u+fG5/16+VJuPPkkQvjk+3yHIIqiCCIiIiKiDsikDkBEREREgYtlkYiIiIg8YlkkIiIiIo9YFomIiIjII5ZFIiIiIvKIZZGIiIiIPGJZJCIiIiKPWBaJiIiIyCOWRSIiIiLyiGWRKIwIgtCtf7799luvfcxvv/3Wa9fMz89vk1MmkyE2NhYLFy7Ejh07+h5WImazGY899liHX6O3334bgiAgPz/f77mIiABAIXUAIvKfMwvVk08+ic2bN2PTpk1tHh8+fLjXPub48eOxY8cOr17zN7/5Da655hq4XC7k5OTg8ccfx+zZs7Fjxw6MGzfOax/HX8xmMx5//HEAwDnnnNPmuQsuuAA7duxAcnKyBMmIiFgWicLKWWed1ebX8fHxkMlk7R73psjISK9fPyMjo/Wa06ZNw8CBAzF37ly8/PLLeO211zp8j8VigUajgSAIXs3SF6Iowmq1dvqa+Ph4xMfH+ykREVF7vA1NRG3U1tZi2bJlSE1NhUqlwoABA/Dwww/DZrO1eZ0gCLjrrrvw3//+F4MHD4Zarcbw4cOxYsWKNq/zdBt6165dWLRoEWJjY6HRaJCVlYV77723V5lPF8eCggIAP9+63bBhA2666SbEx8dDp9PBZrPB7Xbj73//O4YOHQq1Wo2EhARcf/31KC4ubnPNc845ByNHjsSWLVtw1llnQavVIjU1FY888ghcLlefvmb/+c9/MGzYMKjVarzzzjutZfDxxx9vvcW+ZMmSNp/Lmbeh33zzTYwZMwYajQYxMTG49NJLkZub2+Y1S5YsgcFgwIkTJ7Bw4UIYDAakp6fjgQceaJftlVdewZgxY2AwGBAREYGhQ4fiD3/4Q8//ZxBRyOHIIhG1slqtmD17NvLy8vD4449j9OjR2LJlC5566ikcOHAAn3/+eZvXf/rpp9i8eTOeeOIJ6PV6vPzyy1i8eDEUCgWuuOIKjx/nq6++wqJFizBs2DD885//REZGBvLz87Fhw4Ze5T5x4gQAtBuBu+mmm3DBBRfgvffeQ3NzM5RKJe644w68+uqruOuuu3DhhRciPz8fjzzyCL799lvs27cPcXFxre8vLy/H1Vdfjd///vd44okn8Pnnn+PPf/4z6urq8O9//7tXX7M1a9Zgy5YtePTRR5GUlISYmBh8+eWXOP/883HzzTdj6dKlHX4uv/TUU0/hD3/4AxYvXoynnnoKNTU1eOyxxzB16lTs2bMHgwYNan2tw+HARRddhJtvvhkPPPAAvv/+ezz55JMwGo149NFHAQArVqzAsmXL8Jvf/AbPPvssZDIZTpw4gcOHD/fq/wcRhRiRiMLWDTfcIOr1+tZf/+c//xEBiB9++GGb1z399NMiAHHDhg2tjwEQtVqtWF5e3vqY0+kUhw4dKg4cOLD1sc2bN4sAxM2bN7c+lpWVJWZlZYkWi6VHeU+dOiUCEJ9++mnR4XCIVqtV3Lt3rzhp0iQRgPj555+LoiiKb731lghAvP7669u8Pzc3VwQgLlu2rM3ju3btEgGIf/jDH1ofmzVrlghAXLt2bZvX3nLLLaJMJhMLCgpEUez518xoNIq1tbVtXltVVSUCEP/0pz+1+5xPfy6nTp0SRVEU6+rqRK1WKy5cuLDN6woLC0W1Wi1ec801rY/dcMMNHWZbuHChOGTIkNZf33XXXWJUVFS7j01EJIqiyNvQRNRq06ZN0Ov17UYFT98S/eabb9o8PnfuXCQmJrb+Wi6X46qrrsKJEyfa3dY97dixY8jLy8PNN98MjUbTq5y/+93voFQqodFoMGHCBBQWFuK///0vFi5c2OZ1l19+eZtfb968uc3nc9rkyZMxbNiwdp9fREQELrroojaPXXPNNXC73fj+++8B9PxrNmfOHERHR3f/kz3Djh07YLFY2n0O6enpmDNnTruPJwgCFi1a1Oax0aNHt96yB1o+//r6eixevBhr165FdXV1r/MRUehhWSSiVjU1NUhKSmq3CCQhIQEKhQI1NTVtHk9KSmp3jdOPnfna06qqqgAAaWlpvc55zz33YM+ePdi7dy/y8vJQVlaGW2+9td3rzlxBfDpTRyuLU1JS2mX+ZRE+7czPr6dfs76uau7p56DT6dqVcrVa3WZhzXXXXYc333wTBQUFuPzyy5GQkIApU6bg66+/7lNWIgoNLItE1Co2NhYVFRUQRbHN45WVlXA6nW3m8wEtc/rOdPqx2NjYDj/G6bl4nkYeuyMtLQ0TJ07E+PHjMWDAAI8rnM98/HSmsrKydq8tLS1t9/lVVFS0e92Zn19Pv2Z9XY3d08+hu2688UZs374dDQ0N+PzzzyGKIi688MI2I5BEFJ5YFomo1dy5c2EymbBmzZo2j7/77rutz//SN99806ZQuVwurFy5EllZWR5HDgcPHoysrCy8+eab7Vbk+tqcOXMAAP/73//aPL5nzx7k5ua2+/yamprw6aeftnls+fLlkMlkmDlzJoCef806olarAbRs79OVqVOnQqvVtvsciouLsWnTpm59vM7o9XosWLAADz/8MOx2O3Jycvp0PSIKflwNTUStrr/+erz00ku44YYbkJ+fj1GjRmHr1q3461//ioULF2LevHltXh8XF4c5c+bgkUceaV0NfeTIkXbb55zppZdewqJFi3DWWWfhvvvuQ0ZGBgoLC/HVV1/h/fff99nnN2TIENx666148cUXIZPJsGDBgtbV0Onp6bjvvvvavD42NhZ33HEHCgsLMXjwYKxfvx6vvfYa7rjjDmRkZADo+desIxEREejXrx/Wrl2LuXPnIiYmBnFxcejfv3+710ZFReGRRx7BH/7wB1x//fVYvHgxampq8Pjjj0Oj0eBPf/pTj78ut9xyC7RaLaZNm4bk5GSUl5fjqaeegtFoxKRJk3p8PSIKMRIvsCEiCZ25GloURbGmpka8/fbbxeTkZFGhUIj9+vUTH3roIdFqtbZ5HQDxzjvvFF9++WUxKytLVCqV4tChQ8X333+/zes6Wg0tiqK4Y8cOccGCBaLRaBTVarWYlZUl3nfffZ3mPb0a+plnnun0dadXEO/Zs6fdcy6XS3z66afFwYMHi0qlUoyLixOvvfZasaioqM3rZs2aJY4YMUL89ttvxYkTJ4pqtVpMTk4W//CHP4gOh6NPX7OObNy4URw3bpyoVqtFAOINN9zQ5nM5vRr6tNdff10cPXq0qFKpRKPRKF588cViTk5Om9d09P9XFEXxT3/6k/jLb//vvPOOOHv2bDExMVFUqVRiSkqKeOWVV4rZ2dkdZiWi8CKI4hkTbYiIukEQBNx5552t+w2GmnPOOQfV1dU4dOiQ1FGIiCTFOYtERERE5BHLIhERERF5xNvQREREROQRRxaJiIiIyCOWRSIiIiLyiGWRiIiIiDxiWSQiIiIij1gWiYiIiMgjlkUiIiIi8ohlkYiIiIg8YlkkIiIiIo9YFomIiIjII5ZFIiIiIvKIZZGIiIiIPGJZJCIiIiKPWBaJiIiIyCOWRSIiIiLyiGWRiIiIiDxiWSQiIiIij1gWiYiIiMgjlkUiIiIi8ohlkYiIiIg8YlkkIiIiIo9YFomIiIjII5ZFIiIiIvKIZZGIiIiIPGJZJCIiIiKPWBaJiIiIyCOWRSIiIiLyiGWRiIiIiDxiWSQiIiIij1gWiYiIiMgjlkUiIiIi8ohlkYiIiIg8YlkkIiIiIo9YFomIiIjII5ZFIiIiIvKIZZGIiIiIPGJZJCIiIiKPWBaJiIiIyCOWRSIiIiLyiGWRiIiIiDxiWSQiIiIij1gWiYiIiMgjlkUiIiIi8ohlkYiIiIg8YlkkIiIiIo9YFomIiIjII5ZFIiIiIvKIZZGIiIiIPGJZJCIiIiKPWBaJiIiIyCOWRSIiIiLyiGWRiIiIiDxiWSQiIiIij1gWiYiIiMgjlkUiIiIi8ohlkYiIiIg8YlkkIiIiIo9YFomIiIjII5ZFIiIiIvKIZZGIiIiIPGJZJCIiIiKPWBaJiIiIyCOWRSIiIiLyiGWRiIiIiDxiWSQiIiIij1gWiYiIiMgjhdQBiIi8zepwwWx3wWx3wmI//d8uWBxOmO0u2J1uuNwiRACiKMItAhMbCxBfXwHI5BDkMkAmB2QCBLkckMkgKJSQGfSQG42QR0ZCHhkJmV4v9adKRORzLItEFNCabU5Um2yoabajxmRHzU//XW2ytfy6ueXftc12NNucsDhccIs9/zj/q9sI53df9uxNSiXkEREtxdEYCXlE5M//HflTqTRGQhYZCbkxCsqkRChTUiAolT0PSEQkEZZFIpKU1eFCcZ0ZRbUWFNWZUVTb9r8brU6pI3rmcMBVWwtXbW333yOXQ5mYCGV6OpTpaVCl/fTv9HQo09OhiI72XV4iol5gWSQivyiuM+NYRROOlptwvKIJ+TXNKKqzoKrJJnU0/3K54CgthaO0FNi1q93TMoMByrQ0qNLToDxdJNPSoMrIgDIjA4KMU82JyL9YFonIq6qabD+VwqaWf1c04USFCU22AB4hDCBukwm2I0dgO3Kk3XOCTgfN0KHQjBgBzYjh0I4YAVVWFgskEfkUyyIR9Vqj1YEfi+pxoLAe+4vqkV1cj2qTXepYIUs0m2HZtw+WfftaH+uwQA4Y0LIwh4jIC1gWiahbXG4RR8obsb+wHgeKWv7JqzJB7MVikkDkdDikjtArHRZIrbalQI4cyQIZJPLz85GZmYn9+/dj7NixUschakMQxVD5Vk9E3mR3urGvsA7bT1Rj16laHCxpgNnukjqWz/xtz38wpuSE1DF8RqbTQTdpEvRnT4X+7LOhHjRI6khBRRCETp+/4YYb8Pbbb/f6+i6XC1VVVYiLi4NC0bdxnJqaGowZMwYlJSWoq6tDVFRUn65HxJFFIgIAuN0ickobsS2vGttOVOOH/DpYHKFbDs8kV4T2qJvbbIbpu+9g+u47AIAiIQH6qVOhn3Y29FOnQhEfL3HCwFZWVtb63ytXrsSjjz6Ko0ePtj6m1Wr7dH25XI6kpKQ+XeO0m2++GaNHj0ZJSYlXrkfEWdFEYSyvyoR3d+Tj9vf2YtyTX2PRv7fib18cwZbj1WFVFAFAJg+vn52dlZVoWLsWpf/3OxyfMRMnL7oYFX97GqYtW+C2WKSOF3CSkpJa/zEajRAEoc1jy5cvR1ZWFlQqFYYMGYL33nuvzfsFQcArr7yCBQsWQKvVIjMzE6tWrWp9Pj8/H4Ig4MCBA62P5eTk4IILLkBkZCQiIiIwY8YM5OXldZrzlVdeQX19PX7729969fOn8BZe3x2JwpzT5cbu/FpsPFyJjbkVKKw1Sx2JAoTt2DHYjh1D7dtvQ1CpoB03Dvqzz4b+7KnQjBjBFdedWL16Ne655x48//zzmDdvHj777DPceOONSEtLw+zZs1tf98gjj+Bvf/sbXnjhBbz33ntYvHgxRo4ciWHDhrW7ZklJCWbOnIlzzjkHmzZtQmRkJLZt2wan0/OuAocPH8YTTzyBXbt24eTJkz75XCk8sSwShTiTzYnvjlbh68Pl+PZYFerNwbmQg/xHtNth3rUL5l27UPXcc5AbjdDPnInIBefDMH06BJVK6ogB5dlnn8WSJUuwbNkyAMD999+PnTt34tlnn21TFn/1q19h6dKlAIAnn3wSX3/9NV588UW8/PLL7a750ksvwWg0YsWKFVD+dOLP4MGDPWaw2WxYvHgxnnnmGWRkZLAsklexLBKFoLIGCzYersDXuZXYmVcDu8stdSQKYq6GBjSuW4fGdesgi4xExNy5iFy4APqpUyH0cTFGKMjNzcWtt97a5rFp06bhhRdeaPPY1KlT2/36l7edf+nAgQOYMWNGa1HsykMPPYRhw4bh2muv7X5wom7in3KiEFHbbMfn2aVYc6AU+wrrQmZLGwos7sZGNKxejYbVqyGPikLE/PmIXLgAusmTw3prnjNXS4ui2OUK6o7ed1pPF8xs2rQJBw8exEcffdT68QEgLi4ODz/8MB5//PEeXY/ol1gWiYKY2e7EhpwKrD1Qgi3Hq+F0syGS/7jq61G/ahXqV62CPC4Okeeei8iFC6CdMKFbRSlUDBs2DFu3bsX111/f+tj27dvbzUXcuXNnm9fs3LkT48aN6/Cao0ePxjvvvAOHw9Gt0cWPP/4Yll8sTNqzZw9uuukmbNmyBVlZWT39lIjaYFkkCjJOlxvfH6/Cmv2l2JhbEdJ7H1LwcFVXo275ctQtXw5FYiIizz8PkQsWQBsGG0w/+OCDuPLKKzF+/HjMnTsX69atwyeffIKNGze2ed2qVaswceJETJ8+He+//z52796NN954o8Nr3nXXXXjxxRdx9dVX46GHHoLRaMTOnTsxefJkDBkypN3rzyyE1dXVAFqKLPdZpL5iWSQKEodKGrByTxE+P1iG2mYeqUeBy1lRgdp33kXtO+9CmZKCyAsuQNRVV0GVlip1NJ+45JJL8MILL+CZZ57B3XffjczMTLz11ls455xz2rzu8ccfx4oVK7Bs2TIkJSXh/fffx/Dhwzu8ZmxsLDZt2oQHH3wQs2bNglwux9ixYzFt2jQ/fEZEbfEEF6IAZrY7se7HUizfVYgfixukjhPS/pH9FoafzJE6RuiSyWCYMQPR1yyGfsaMsNuKRxAErF69GpdcconUUYh6jCOLRAHocGkjlu8uwNr9pWiyed5XjShouN2tJ8go09MRffVVMF52GRTR0VInI6IusCwSBQiL3YV12S2jiAeK6qWOQ+QzjqIiVD7zLKr+9SIiFyxA9K+vgXbUKKljEZEHLItEEiuqNePNbafw0d5iNFk5ikjhQ7TZ0LBmDRrWrIFm5EhEL16MyAsvgEytljqa13HGFwUzzlkkksiBonq89v1JfJlTDhe3vJEc5ywGBrnRCONllyF68dVQZWRIHYeIwLJI5FeiKOLrwxV4fcsp7M6vlToO/QLLYoARBOinT0fM9dfBMGOG1GmIwhpvQxP5gdXhwsf7ivHG1lM4WdUsdRyiwCeKaN6yBc1btkAzahTi7rgDEXNmd/0+IvI6lkUiH2owO/D29ny8uyMfNdwbkahXrAcPonjZMmiGD0fcnctgmDMnrE6IIZIayyKRDzRYHHhjy0m8tS2fW98QeYn18GEU33kX1MOGIe6O2xExfz5LI5EfsCwSeVGj1YE3t57CG1tPcWUzkY/YcnNRcvc9UA8Z0nJ7+rxzWRqJfIhlkcgLmqwOvLUtH69vOYlGlkQiv7AdPYqSe++FetDAltJ4/vlhdzIMkT9wNTRRH5hsTry97RRe23IKDRaH1HGoD7gaOvipsrIQd/vtiLxgIUsjkRexLBL1gs3pwjvb8/HKt3moM7MkhgKWxdChysxE3B23I/LCC1kaibyAZZGoh9b9WIq/f3UERbUWqaOQF7Eshh710KFI/P3voD/rLKmjEAU1zlkk6qa9BbX48+e52F9YL3UUIuoG25EjKFxyIwyzZyPh/x6EOjNT6khEQYllkagLhTVm/O3LXKw/WC51FCLqBdPmzTBt3YroxVcj/s47ITcapY5EFFRYFok8aLA48OI3x/HujgLYXW6p4xBRXzgcqHv3PTSu/RRxdy5D9DXXQFDwr0Ci7uDMX6IzuN0i3t2Rj1nPbMbrW0+xKBKFEFdDAyr++hROXnIJmnfulDoOUVDgj1VEv3CwuAEPrzmI7OIGqaMQkQ/ZT+ShcMmNiDj/fCT+/ndQJiVJHYkoYLEsEqFlU+1nvzqK93YWwM39AYjCRtOXX8L0/feIu+02xN64BIJKJXUkooDD29AU9j79sRRz//Ed3tnBokgUjkSzGVXPPYeTiy6CacsWqeMQBRyWRQpb+dXNuO6NXbj7g/2obLJJHYeIJGYvKEDRLbei5P4H4KqvlzoOUcBgWaSwY3O68NzXx3Du899jy/FqqeMQUYBpXL8eeYsWoWnTJqmjEAUElkUKK9nF9bjwX1vxwjfHYXdylTMRdcxVVY3iZXei9He/g6uxUeo4RJJiWaSw4HC58Y8NR3HZy9txvNIkdRwiChINaz/FyQsXwfTdd1JHIZIMyyKFvMOljbjo39vw4qYTcHIFCxH1kLOyEkW33Y7SPzwMV1OT1HGI/I5lkUKW0+XGi98cx8UvbUVuGW8jEVHfNHzyCU5edDFMW7dJHYXIr1gWKSQdr2jCZa9sxz++PgaHi6OJROQdzrIyFC1dirJHHoXL1Cx1HCK/YFmkkOJ2i3j1+zxc8OJWnsJCRD5Tv2oVTl60CM07dkgdhcjnWBYpZFQ2WXHdm7vw1/VHuNKZiHzOWVqGwptuRtljj8HdzFFGCl0sixQSvj9WhYUvbMG2EzVSRyGicCKKqF+xEqcuuxzWo8ekTkPkEyyLFNScLjee/vIIbnhzN6pNdqnjEFGYshcUIP/qq1G/Zo3UUYi8TiF1AKLeKqm34O4P9mNvQZ3UUYiIIFosKPv9Q7Ds24/EPz4MmUoldSQir+DIIgWlDTnlWPD89yyKRBRw6j/8EAWLr4G9uETqKERewbJIQcXudOOxT3Nw63t70Wh1Sh2HiKhD1pwc5F9+OU9+oZDAskhBo6zBgstf2Ya3t+dLHYWIqEuuhgYU3X4HKl94AaKbOzRQ8GJZpKCw62QNFr7wPQ6W8CQWIgoiooiaV/6DoqVL4azjtBkKTiyLFPDe3HoK17y2E3Vm3nYmouDUvH0HTl16GSwHDkgdhajHWBYpYNmcLtz7wT488dlh8MQ+Igp2zvJy5F93PWrffU/qKEQ9wrJIAamy0YpL/70Fa34skzoKEZH3OByo+OtfUXL//Tz1hYIGyyIFnB+L6nDec5txuJzfSIkoNDWu/wL5110HR2Wl1FGIusSySAHlw935uPzl7aizcOUgEYU22+FcFFy9GLa8PKmjEHWKZZECxt8+y8b/fZIDJ+cnElGYcJSWIv+aX8P8ww9SRyHyiGWRJOdyi7j9ra34z9YiqaMQEfmdu6EBhTfdjMYvv5Q6ClGHeDY0SarZ6sDilzcju9IhdRQiIsmIdjtK7rsfjrJyxN64ROo4RG1wZJEkU1zdiHOf+YpFkYgIAEQRlU8/jfK//pUnvlBAYVkkSew9UYqFz3+LkmZB6ihERAGl7t33UHLvfXDbbFJHIQLAskgS+HRnLq554wc0OuVSRyEiCkhNGzag8Mab4KqvlzoKEcsi+dfLn+/BfWtOwCayKBIRdcaybx/yF18De3GJ1FEozLEskl+Iooi/rPwOz3xfDhd/2xERdYv91CnkX301LIdypI5CYYx/a5PPOZ1O/O7NDXhtXxNEgb/liIh6wlVdjYLrr4dpy1apo1CY4t/c5FMWqxW/+c96fHjMAQhczEJE1Bui2YziO+9kYSRJsCySzzQ0mXDLi5/hi2I5iyIRUR+JdjuK77oLpq3bpI5CYYZlkXyivKoGN/7rM2yt0UodhYgoZIg2W8sI4zYWRvIflkXyupMFxbjppS+xrylC6ihERCGnpTDehebt26WOQmGCZZG86vCxPCx9dTMOW6OkjkJEFLJEqxVFy+5E844dUkehMMCySF5z8Mhx3Pn2Npx0xUgdhYgo5IlWK4ruWIbmnTuljkIhjmWRvGLfoVzc9+5WnHLHSh2FiChs/FwYd0kdhUIYyyL12e79B/G797fhhDtB6ihERGFHtFhQdMcdaN61W+ooFKJYFqlPtu3Zj0dWbsdxMVnqKEREYUu0WFB0++1o3s3CSN7Hski9Iooivt3xA/68ajuOIk3qOEREYa+lMN4B8549UkehEMOySD0miiI2bt2Fpz/ejlxZP6njEBHRT0SzGYW33Q7zDz9IHYVCCMsi9Ygoivjqu+14fs0OHFEMAMCTWYiIAoloNqPottthPXJE6igUIlgWqdtEUcSmbbvx8rqdOKwcCJFFkYgoILmbm1F02+1wlJdLHYVCAMsiddvW3fvx37XfIUc5CCJ/6xARBTRnRQWKbr0NLpNJ6igU5Pg3PnXLrv0H8eonG3BANhguyKWOQ0RE3WA7dgwld98D0emUOgoFMZZF6tLeg7l4beVn2IuBsAtKqeMQEVEPNG/fjrJH/yR1DApiLIvUqezcY3h95VrscfeHWdBKHYeIiHqh4ZNPUPXSS1LHoCDFskge5R4/idc/WI0fbMlokEVKHYeIiPqg+sV/o37NGqljUBBiWaQOHT9ViNeWf4J9zdGoksdJHYeIiLyg7JFH0bxzp9QxKMiwLFI7xWUVeH35JzhQr0SxMlXqOERE5C0OB4p/czdsx49LnYSCCMsitVFb34A3VqzGjxVWnFJnSR2HiIi8zN3UhMLbboOjslLqKBQkWBapVbPZgrc+XIu9eWXI04/gpttERCHKWVqG4tvvgNtsljoKBQGWRQIAOBxOvL96PbYfyEV+5Fg4RP7WICIKZdbDh1Fy3/0QXS6po1CAYyMguN1urP7yG2zasQdlMWPR6FZJHYmIiPzA9N13qPz736WOQQGOZTHMiaKIr7fsxLqN36MuaihKnQapIxERkR/VvvMuGr/8SuoYFMBYFsPc7gOHsOqzr2HWJ+Ook1vkEBGFo7I//hH2/HypY1CAYlkMY7nHT+Ldj9ehSVQjW+wHcEELEVFYcptMKL7nXritVqmjUABiWQxTZZXVeOejdahtsiJXPYwLWoiIwpzt6FGUP/mk1DEoALEhhKFmswXvfrQOp4pKUWgcjUaXQupIREQUABo+/gT1n6yWOgYFGJbFMONyubDqs6+x/1Au7EkjUWLXSh2JiIgCSPkTT8B69JjUMSiAsCyGmW+27cbGrTuhS+yPA+ZoqeMQEVGAEa1WlNxzD1ymZqmjUIBgWQwjPx4+ho8//xpagxG77ak8oYWIiDpkz89H2R//KHUMChAsi2GipLwS76/+HFa7A8eUWTBxniIREXWi6csvUfve/6SOQQGAZTEMmJrNePejdSgqLYctbggKbJynSEREXav8+99hyc6WOgZJjGUxxLlcLqz89Cv8ePgoYtIHYk9TlNSRiIgoSIgOB4rvvReu+nqpo5CEWBZD3Kbte7B5xx4kp6Rga1MCXJynSEREPeAsLUPJ734HURSljkISYVkMYcdPFeKTL75BhEGPQ65UNHCeIhER9ULzd9+j7t13pY5BEmFZDFGNTSa8v2Y9mpqaYYtMx3GLTupIREQUxCqffwH2ggKpY5AEWBZDkNvtxkfrN+LI8ZNIyeiH7Y1RUkciIqIgJ1osKH34Yd6ODkMsiyFoy+592Lx9D9JTkrCrORZWt1zqSEREFAIsP+xF3f/elzoG+RnLYojJLyrFR59vhFajRpUiAQVWbpNDRETeU/ncc7AXFUkdg/yIZTGENJsteH/1elTX1iM2MQU7GoxSRyIiohAjms0oe/iPvB0dRlgWQ4Qoilj95SYcPHIMAzPTsbUhGjaR/3uJiMj7zLt3o275cqljkJ+wTYSI3QcO4estO5GSlIACpxFFNo3UkYiIKIRV/uOfsBcXSx2D/IBlMQRU19bj4/UbIZMJUEVEYydvPxMRkY+JZjPK/vgIb0eHAe7SHOTcbjc+/mIjCorLMHxwFr6uj4Kdt5+J6Cc/mM14s7YGOVYbqlxO/CslFfMiItq8Js9mwz+rqrDHYoZbBAaqVfhnSipSlMoOr3lDYQH2WCztHp+p1+M/aekAgHWNDXiuqgpmtxuXG6PwYEJC6+tKHHYsLSrCqn79YZBzt4ZgZt65E/UrVyL66quljkI+xLIY5Hbs/RFbd+9Hv/QU5Nv0KObtZyL6BbPbjSFqDS41RuGe0pJ2zxfa7bi2sACXG6NwZ1wcImQynLTboRY8Hw36QmoaHL8YTap3uXBZ/imc91MJrXM68Wh5Of6alIw0pRJ3lBRjsk6HWQYDAODxigrcH5/AohgiKp95FoYZM6BMTZU6CvkIy2IQq6iuwSdfbIJKqYRGZ8DOykipIxFRgJlpMGDmTyWtIy9UV2GmwYDf/mLkL12l6vSaUWeUvC8aG6GRyXBeRMv3oCKHAwaZDAsiW349WafDCbsNs2DAZ40NUAoC5p8xuknBy93cjLJHHkHGm29KHYV8hPcrg5TL5cLH679BcXkF+qUl44emCFi4+TYR9YBbFPGdqRn9lSrcUlSE6SeO46qCfGxsaurRdT5uaMDCiAjoZC1/pfRTqWAVRRy2WlHvcuGQ1YohajXqXS68WF2NPyYk+uLTIQk1b9+BupUfSh2DfIRlMUht++EAdvxwAP3TUlDjVCO3WS91JCIKMjUuF8yiG6/X1mC6Xo/X0tIxzxCBe0pLsMds7tY1si0WHLfbcLkxqvUxo1yOp5KS8VBZGa4qyMdFkZGYrjfgmcpKXBsdjRKHA5fln8JFp07iq6ZGH3125G+Vf/87HOXlUscgH+Bt6CBUVlmN1V9uglqjhkGvxzfVRojwPL+IiKgjIlrmHc4xROCGmBgAwDCNBgcsFqysr8Mkna7La3zc0IBBKjVGa9ueFjUvIqLNQprd5mYct9vwx8REnH/yJJ5NSUGcQo6rCgowUatDrIJ/HQU7d3MzKv/+d6T+859SRyEv48hikHE6nfh4/UaUVVQjIyUJuWYdqh2dzy8iIupIlFwBBYAsddvvIQPUKpQ5nF2+3+J244umRlwR1fl2XXa3G09UVOCxxCQU2u1wQcQknQ6ZKjX6q1TItrZfWU3BqXH9FzDv2SN1DPIylsUgs2NvNnbuy0b/9BRYRQV+aOSiFiLqHZUgYKRGi1N2e5vH8+12j9vm/NKXTY2wiyIWRXZeFl+pqcEMvR7DNRq4ADh/sZLaIYpwcZu+kFL+579AdLmkjkFexLIYRGrrG7Bu43dQq1Qw6HXY1RjJPRWJqFPNbjdyrVbkWq0AgBKHA7lWK0odDgDATTEx+KKxEavq61Fgt+P9ujp8azLh6qio1mv8vqwU/6yqbHftjxsaMNdgaLc6+peO22z4oqkRv4mLBwAMUKkgEwR8XF+P70wmnLLbMUrDLb9Cie3oUdStWCF1DPIiThIJEqIo4vNvtqCwpAzDBmWh1KZCnqXr+UREFN5yrBYsKSpq/fXTP5W+SyIj8dfkFMyLiMCfkpLwWk0N/lpZgf4qFZ5PScWEX8xXLHM42o0s5Nvt2Gex4PWfNuHuiCiKeKy8HL9PSGxdKa2RyfDXpGQ8WVEOuyjijwmJSOzGKCYFl+p/vYjIhQuhiI6WOgp5gSDynJ6gcPj4Sfzz1XcRYdAjJioKa6riUePkN1gib/lH9lsYfjJH6hhEISPq6quQ/NhjUscgL+A9zCBgs9mx5stNsFhtiI2OwnGLlkWRiIgCWsO6z1BaclTqGOQFLItB4Ltde3Hw6AlkZqTB6Ra4qIWIiAKXIKBxznj8YZkBjxz6u9RpyAs4ZzHAlVVW47ON3yHSoIdGrcL+Jj3MPKmFiIgCkHvEILw1T8BXuuyWB8prsKlwE+ZkzJE2GPUJRxYDmNvtxqcbNqOiuhapSQkwu2TINnk+45WIiEgKQmICvrt5HK6+6BS+0p1s89xze5+Dw+2QKBl5A8tiANufcxTb92YjIzUZMpkMe5si4OBWOUREFCBElQqnrpiMpTea8VLCwQ5fk9+Yj5VHVvo5GXkTm0eAslit+HTDtxBFEcYIA2odChwzc6scIiKSngggNz0Cv71Rgd8N2ocGwdrp61/NfhXNjmb/hCOvY1kMUNv2HMCxUwXon5YCANjVGMnzn4mISHK18UY8tzgaf7rWgqIYe9dvAFBnq8O7h9/1cTLyFZbFAFTX0IgvNm+DQa+DSqVEsVWNEhtPOCAiIulYdVp8dF4Cbr/JhJ39m3r8/ndz3kWDrcEHycjXWBYD0Ddbd6GkvAJpSYkAgL1NERInIiKicOWSy7FjUiJuv92FD8fXArLe3eUyOUx489CbXk5H/sCyGGCKyyqwadtuxMfGQC6XociqRpVDJXUsIiIKMyKAvIHx+N1SLZ6bVwOz2t3na35w5ANUW6r7Ho78imUxgIiiiPWbtqK2vhEJcTEAgH0cVSQiIj87PS/xoV/VoTCm88UrPWFxWvBq9qteux75B8tiADl8/CR27c9GWnIiBEFAIUcViYjIj6w6LT4+t/fzErvjo2MfodRU6pNrk2+wLAYIp9OJ9Zu2wGqzI8rYMprIUUUiIvIHl1yOnROTsOw2N1ZO6P28xO5wuB34z4//8dn1yftYFgPED9mHkX34GPr9tFVOgVWNao4qEhGRj538aV7iP+dXw6Rx+eVjfpr3KQoaC/zysajvWBYDgNVmwxebt0Eul0On1UAUOapIRES+VRdnxPNXx+D3Xp6X2B0u0cWV0UGEZTEA7DmQgxP5hUhPTQIAFFg1qOGoIhER+YBVp8HqcxNw280mbM9slCzHurx1qGiukOzjU/exLErMarNhw/c7oFIqoVapIIrAfo4qEhGRl7nkcuyamIhlt4n4wMfzErvD4XbwVJcgwbIosd0HDiGvsAhpKS0bcBfb1KhxKiVORUREoeRUVjx+v1SLf8yv8du8xO746NhHPNUlCLAsSshitWLD9zugVqmgVrXcdj5oMkicioiIQkVdnBEvXBWD311ZhwI/z0vsDrPTjOW5y6WOQV1gWZTQ7gOHcLKwGOnJLXMVaxwKlNrVEqciIqJgZ9NqsGZ+Im672YRtA6Sbl9gdy48sh9lhljoGdYJlUSIWqxVfb9kJjUoNlarltnM2RxWJiKgPXHI5dk9Iwh23i1g+sUbyeYndUW+rx8fHP5Y6BnWCZVEiu/YfwsmCYqQlt8xVbHbJcNKilTgVEREFq1MD4vHQzTo8e67/9kv0lndy3oHD5ZA6BnmgkDpAOLJYrdi4ZSc0mp9HFQ+ZDBAR+D8BEhFRYKmPM+LduQpsHVAndZReqzBXYP2p9bh44MVSR6EOcGRRAqfnKp4eVbS7BRwx6yRORUREwcSm1WDtvATcerMJWwcE/4ri5Ue40CVQcWTRz+x2BzZt2wO1WgWVsmVU8ahZB4fI3k5ERF1zyeXYNyYOr8yqh0lTC4TIXanDNYdxoPIAxiaMlToKnYFl0c9+zD2GvIIi9E9vOQPaLQI5zXqJUxERUTDIHxCPV+ZZcSq2RuooPvHBkQ9YFgMQh7P8yO124/tdeyEA0KhbtsjJt2pgcrGzExGRZ/WxkXjxylj831V1OBVrkTqOz2wo2IBqS7XUMegMLIt+dOxkAXKO5iElKaH1sVyOKhIRkQc2rQZr5ybg9qVmbMkK/nmJXXG6nVh1bJXUMegMLIt+tGX3PlhtNkQYWgpig1OOMm7CTUREZ3DLZPhhXCLuvE3E+5Nr4Q6jv61XHV0Fh5vb6AQS3v/0k+KyCuzNPozE+NjWx45wVJGIiM5QkBmH/8yzIS8uNOcldqXKUoVvCr7B+ZnnSx2FfsKy6Cc79v6I+kYTUn/aLsclAse4CTcREf2kPjYS789R4ruB9VJHkdzyI8tZFgMIy6If1NY3YMvufYiNiYIgtGxxcMqihc0tlzgZERFJzaZR46tpUVg+sSasbjd3Zn/lfhyrO4bB0YOljkLgnEW/2H3gECqr65AQG9P6GDfhJiIKb26ZDHvHJeLO24H/TWZRPNOaE2ukjkA/4W9NH7PZ7Ph+514YDDrI5S1f7nqHAuVc2EJEFLYK+sfh4Zv1ePr8GjRqg+scZ3/5/OTnXOgSIHgb2scOHctDYWk5MjNSWx/jqCIRUXhqiInE+3OV+JbzErtUa63F98XfY27GXKmjhD2WRR8SRRG79mVDFEWoVSoAgFMEjrMsEhGFFbtGjQ1nR+F/k3i7uSfWnFjDshgAWBZ9qLSiCtm5x5AQ9/N2OYVWDWw8B5qIKCy4ZTLsHxOP/8xqQIM2PLfC6YutxVtRbalGnDZO6ihhja3Fh/YfykV9owkxUZGtj53gdjlERGGhsH8cHrmpZV5ig9YpdZyg5BSd+Pzk51LHCHssiz5is9mx7YcfERGhb90ux+oWUGzVSJyMiIh8qSEmEi9fEYvfLq7H8fjQPcfZX7gqWnq8De0jB4+eQHFpOfr/YmFLvkULNwQJUxERka9wXqJvnKg/gUPVhzAybqTUUcIWy6IPnF7Y4v7FwhaAt6CJiEKRWybDgdHxeOUczkv0lU/zPmVZlBB/9vGBkvJKZB853uYc6GaXDOV2VSfvIiKiYFPUr2Ve4t8WcF6iL20s2Ai36JY6RtjiyKIP7D90BA2NJqT9dA40AORZtABvQRMRhYTG6Agsn6PGpsH1UkcJC1WWKuyt2ItJSZOkjhKWWBa9zOFwYuf+bEQYdK0LWwDgBPdWJCIKenaNGhunRuHdSTVwy7l4xZ825G9gWZQIb0N72Yn8QhSXVSL+F+dA1zkUqHUqJUxFRER90bJfYiLuul3A22fVwC2XOlH42VjIW9FS4ciil2XnHoPdbodO+/MWOXlc2EJEFLSK+sXh1fl2HI3n4hUpVVuqeStaIiyLXmSxWrHnx8OIMka2efwU91YkIgo6jdER+GC2Ct8MqZc6Cv3kq/yvWBYlwNvQXnQkLx/lVdWIi4lqfazBKUcDb0ETEQUNu1qNL85JxK23WPDNkAap49AvcFW0NDiy6EU/5hyD2+1us7diAUcViYiCglsmQ/aoeLw8uwH13C8xINVYa3grWgIsi17S2GTC/pxcxEQb2zzOskhEFPiKM+Lw33M5LzEYcFW0/7Esesnh4ydRXVuPQZkZrY9ZXTJUciNuIqKA1RgdgRWz1djIeYlBY0vJFqkjhB2WRS/ZdzAXAKBQ/PwlLbSpIXIjbiKigGNXq7HprCi8PYX7JQabElMJ8urzkBWVJXWUsMGy6AXVtfU4dCwP8bHRbR7nLWgiosDilslwcGTLvMQ6HW85B6vvi79nWfQjrob2gmMn81Hf0Iho48/zFV0iUGJTS5iKiIh+qSQjFo8tMeAvF9SgTsdznIMZb0X7F0cWvSDnWB4EQYBc/nP3LrWp4RTZxYmIpNYUFYEVc9T4mvMSQ8b+iv1osjchQhUhdZSwwDbTR81mCw4dPYHoqLYbcfMWNBGRtBxqFb6clYhbbrWwKIYYp+jE9tLtUscIGyyLfXT8VCFq6hoQE9V2y5xi3oImIpKEWybDj6MS8Zvb5XjzbJ7jHKq2FPNWtL/wNnQfHc07BZfbBZXy51NaGp1ymFz80hIR+VtJeiz+O9+BI4lcvBLqtpZshSiKEATuOuJrbDR94HA4sT/nKCINhjaPl3JUkYjIr5qiDFg5W4MNQ+uljkJ+UmOtweGawxgRN0LqKCEvYG5D5+fnQxAEHDhwQOoo3ZZfXIqKqhrERke1eZxlkYjIPxxqFTbMTMRtt1pZFMPQzrKdUkcIC90ui4IgdPrPkiVL+hQkPT0dZWVlGDlyZK+vcc8992DChAlQq9UYO3Zsn/J0x7GTBbBabdBqfi6HogiU8tQWIiKfcgsCsn+al/j6tBo4OS8xLP1Q8YPUEcJCt29Dl5WVtf73ypUr8eijj+Lo0aOtj2m12j4FkcvlSEpK6tM1RFHETTfdhF27diE7O7tP1+rOx9p3KBdanabNfIk6pwJWzqYmIvKZ0vRY/HeeE7lJnJcY7vZX7ofL7YJcxr93fanbI4tJSUmt/xiNRgiC0Oax5cuXIysrCyqVCkOGDMF7773X5v2CIOCVV17BggULoNVqkZmZiVWrVrU+39Ft6JycHFxwwQWIjIxEREQEZsyYgby8PI8Z//Wvf+HOO+/EgAEDevAl6J3SiioUlZYj9oxV0LwFTUTkG01RBrxxSRzuvbYBuUnNUsehANDsaMbhmsNSxwh5XpmzuHr1atxzzz144IEHcOjQIdx222248cYbsXnz5jave+SRR3D55Zfjxx9/xLXXXovFixcjNze3w2uWlJRg5syZ0Gg02LRpE/bu3YubbroJTmdg7Lp/qrAETaZmREZwcQsRkS85VCp8/dO8xK+G1UsdhwLMnoo9UkcIeV5ZDf3ss89iyZIlWLZsGQDg/vvvx86dO/Hss89i9uzZra/71a9+haVLlwIAnnzySXz99dd48cUX8fLLL7e75ksvvQSj0YgVK1ZA+dO2NIMHD/ZGXK84kV8EmUzW5ha0WwTKOV+RiMgr3IKAnBEJeHlOI2r0vOVMHdtTvgc3jbxJ6hghzSsji7m5uZg2bVqbx6ZNm9Zu1HDq1Kntfu1pZPHAgQOYMWNGa1EMJA6HE4eP57UbVax2KGHnEX9ERH1WlhaLJ2+IxJOLalCjd0gdhwLY6XmL5Dte22fxzE0xu7tRpqfX9HXBjC8Vl1WgurYeCfExbR7nqCIRUd+YjAZ8OFuLL4fVSR2FgsTpeYuj4kdJHSVkeWUYbNiwYdi6dWubx7Zv345hw4a1eWznzp3tfj106NAOrzl69Ghs2bIFDkfg/USZX1wKi9UK/RmFtoJlkYioVxwqFb6ekYhbb7OyKFKPcd6ib3llZPHBBx/ElVdeifHjx2Pu3LlYt24dPvnkE2zcuLHN61atWoWJEydi+vTpeP/997F792688cYbHV7zrrvuwosvvoirr74aDz30EIxGI3bu3InJkydjyJAhHb7nxIkTMJlMKC8vh8ViaV1ZPXz4cKhU3ityx08VQC6XtxsVrWRZJCLqEbcg4PCIBLw8uxHVBs5LpN7ZX7lf6gghzStl8ZJLLsELL7yAZ555BnfffTcyMzPx1ltv4ZxzzmnzuscffxwrVqzAsmXLkJSUhPfffx/Dhw/v8JqxsbHYtGkTHnzwQcyaNQtyuRxjx45tNzfyl5YuXYrvvvuu9dfjxo0DAJw6dQr9+/fv8+cJAFabDbknTsEYGdHm8SanHBbur0hE1G1lqTF4db4LOcksidQ3OdU5UkcIaYIoiqJfPpAgYPXq1bjkkkv88eF85tjJAvz1xdeRmpwArUbT+nieWYvN9dESJiOivvhH9lsYfpJ/4fiDKVKPVbN1+GI4bzeT92y8YiMS9YlSxwhJXlvgEi7yi0thtduhUbfdT7HSEXirtomIAolDpcT3k2Pw5tRaOBQ2qeNQiDlUc4hl0UdYFnvo6IlTUCmV7eYrVnG+IhFRh0RBwOHhCXhpDuclku8cqj6EuRlzpY4RkvxWFv10t9unrDYbThQUwXjG/opuEahxsncTEZ2p/Kd5iYc4L5F87FD1IakjhCw2nB4oKa9Eo6kZyQnxbR6vcyrg4mbcREStTJF6fHyOFp+PqJc6CoWJnJqcbu/xTD3DstgDJeWVsFpt0GrazlfkLWgiohYOlRJbJsfgDc5LJD9rsjehsKkQ/SL7SR0l5LAs9kBhSTkEQWj3U0s1F7cQUZg7PS/x5dlNqIrgLWeSxsHqgyyLPsCy2E2iKOJo3ikY9Lp2z9WyLBJRGKtIicFr813ITmFJJGnlVOfgwgEXSh0j5LAsdlNNXQOqa+sRYdC3e66ei1uIKAw1R+jx8Tk6fDaS+yVSYDhRf0LqCCGJLaebSisq0WRqRmJ8bJvHTS4Z7FzcQkRhxKn8ab/Es2th57xECiB59XlSRwhJLIvdVFZZDbcoQqFo+yWr4y1oIgoToiAgd1g8Xp5jQiXnJVIAqrJUocHWAKPaKHWUkMKy2E0FxaWQydqPINbxFjQRhYGf5yXWSh2FqFN59XkYnzhe6hghhU2nG1wuF/Lyizucr8iRRSIKZZyXSMEmr4Fl0dtYFruhuq4e9Y2NiIyMaPccRxaJKBQ5lUpsmRSLN6bVcF4iBRXOW/Q+Np1uqKqpQ7PFiuTEtie3iCJXQhNRaBEFAUeGtsxLrIisljoOUY9xRbT3sel0Q1VNHVwud7vFLU0uOZxcCU1EIaIyORqvzxNxII3zEil4cWTR+1gWu6GiugYdHTXJW9De0bDjQ9R//y4iJlyEmHm3AgDqt76P5twtcDVVQZApoEoaiKiZ10OdMsTjdcxHt6Nh54dw1JUBbicU0SmInHQpDCPntL7GlLMZ9d+9A9FhhWH0uYiefVPrc86GClSsfATJNzwPmbr95utEocococcns3T4dBTnJVLwq7ZUc0W0l7HtdENhSTk0Z5wHDQCNLIt9Zis7hqYfv4Iyvn+bx5UxqYiZfzsUUUkQHTY0/bAWFSsfQeptr0Gu6/gbgExrgHHqlVDGpANyBSx5u1Gz/nnIdUZoB0yAy9yA2i9fROzCe6GISkLlR49DnTEKuqxJAICar15G9KwlLIoUNpxKJbZOisHr07hfIoWW/MZ8jIkfI3WMkMF7qF1wOJwoq6iETqtp91yTSy5BotDhtltQve5ZxJ7/G8g0hjbP6YefA23/sVBGJUEV3w/Rc5ZCtJthrzzl8XqajNHQDT4byrh0KKOTETnxYqgSMmErPgwAcNaXQ1DroB82E+rkwdBkjIajuhAA0Hz4WwhyBXRDzvbdJ0wUIFr2S0zAfbcq8fKsGtgVotSRiLyqpKlE6gghhWWxC1W1dTCZLdBrte2ea+LIYp/Ufv0KtFmToO0/ttPXiS4Hmg58CUGthyohs1vXFkURlvwDcNQWQ50+EgCgiEmF6LDBXpEHl6UJ9rJjUMX3h8vShPot7yNm/u19/ZSIAl5lcjSeutaIP11Si4pIu9RxiHyixMSy6E1sO12oqqmD2WJFWnJiu+dMHFnstebD38FenofkG57z+Brzid2o/vTvEB02yA3RSLzqSY+3oE9z25pR/NINEF0OQJAh9tw7oM0cBwCQawyIu+A+VH/2T4hOO/Qj50A7YAKq1z+PiAkXwtlQgcqPnwTcThinXQP90Ole/ZyJpGSO0GP1LB3Wcl4ihQGWRe9iWexCdW0dRFGEXN6+GLIs9o6zsQq137yGxKuegKBQeXydJmM0km/8F9zmRjT9+BWq1j6N5Ov+Abk+yuN7BJUWyTf+C6LdCmvBAdRuegOKqCRoMkYDAHSDz4Zu8M+3mq2F2XBUFSBm/u0offVWxC16EHJ9NMrevR+a9JGdfiyiYOBUKrBtYhxen1YDm5LzEik8sCx6F8tiFyqqOz7/1OqSwcFtc3rFXn4CbnM9yt6+9+cHRTdsRTlo2vcZMn67GoJMDplKA5kqBYhOgTp1KEpevQWm7A0wTr3S47UFQQZldAoAQJU4AI6aYjTsWNVaFn9JdDpQu+EVxF74AJx1ZRDdLmgyRgFoWWBjKzsK3cApXv3cifxFFAQcHRKPl+eaUM79EinMsCx6F8tiFwpKyqDVcHGLN2n6jUHyTf9u81jN+hegjE1D5JTLIcg8fG3FlvmLPSGKosf31G9fAc2ACVAnDYS9Ig9wu35+n9sJuN09+lhEgaIqKRpvzBOxL537JVJ4Km8uh1t0QyZwUMcbWBY74XK5UFVdC20H2+awLPaeTK2D6oytcgSlGjJNBFTx/eG2W9GwYyV0A6dAboiBy9II0/71cDZVQzfk53mE1Z/9A/KIWETPWgKgZb9GVdIgKKKTAZcDlrwf0JyzCTHnLmuXwV5VAPOR75G85EUAgCImDRBkaPpxA+SGaDhqiqFKHuSzrwGRL5gNOqyZZcCa0SyJFN4cbgcqzZVI0idJHSUksCx2oqHJBIvVBp2OK6H9SZDJ4KgtRtWab+CyNEKujYQqaRCSfv00VPH9Wl/nbKwCfvFTo9thQ+3XL8PVVANBoYIyJg1xFz4A/bCZba4viiJqv/o3oufcApmqZdRYplQjduG9qP36FYguR8sejxFx/vmEifrIqVRg+8RYvDatFjYliyIRABQ3FbMseokgiiI32PLgVGEJnnj+v0hJSmg3urit3ohcs16iZETkbf/IfgvDT+ZIHaNHREHAsSEt5ziXGbkNDtEv/WX6X3BR1kVSxwgJHB7rRH1TE2x2O9QqZbvnuBKaiKTEeYlEnau2cGGXt7AsdqK+oQkAIJO1nyBrdXPSLBH5n8Wgw5qZBqwew5JI1JkaS8e7mVDPsSx2oqGxyeNzFpZFIvIjp0KBHRPj8Pr0Wlg4L5GoS7VW/jnxFpbFTlTV1nW4GTfAkUUi8g8RwPEhCXhprgllRt5WI+oujix6D8tiJ8oqq6FWtz9hxCkCTm7ITUQ+Vp0YjTfnAz9wXiJRj3Fk0XtYFj1wOJyorWuAVt1+j0ULF7cQkQ9ZDDqsnaHHJ2N5jjNRb9VYObLoLSyLHjQ0mWC122HQ69o9x1vQROQLLoUCOybE4bUZtbAoWRSJ+qLeWg9RFCEIgtRRgh7LogdNpmbY7HbERhvbPceySETeJAI4PjgBL89rRinnJRJ5hVN0ot5Wj2hNtNRRgh7LogdmqxUOhxNKZfs9FlkWichbqhOj8OY8AT9kcH4VkbfVWmtZFr2AZdGDZrMFbrcb8g72WOS2OUTUVxa9Fp/OjMDHY1kSiXylzsrpHN7AsuiB2WIFgA7nOnBkkYh6y6VQYOeEOLw6g/slEvma2WmWOkJIYFn04HRZ7IiDZZGIekgEcGJwPF6Za0ZxFOclEvmD2cGy6A0six6YLRaIHp5zilxZRUTdV5MQhbfmC9idwVtiRP7U7GiWOkJIYFn0oL6xqcP5igDLIhF1j0WvxboZEfhoHG83E0mBt6G9g2XRg4YmE1QdrIQGWBaJqHMuhQK7xsfiv9PrYFGzKBJJhSOL3sGy6EFDowlKZcdfHgfLIhF1QASQNygeL88zoziKp0cQSY0ji97BstgBl8uFJlMzlIqOvzwulkUiOkNNQhTemifD7n6cl0gUKLjAxTtYFjtgsdrgcDqh8FAWObJIRKdZ9Fp8Nj0Cq8bzdjNRoGFZ9A6WxQ7YHQ64XC5o1OoOn+ecRSJyKeTYNS4er86ohZnzEokCEm9DewfLYgccTifcbjdkcq6GJqK2RAAnB8bj5XkWFEVzv0SiQGZ32aWOEBJYFjvgcDjhcrshk3VcClkWicJTbbwRb8+TYWd/zkskCgZuuKWOEBJYFjvgdLrg8nAuNAC4WRaJwopVp8VnMyLw4dgawMMPkUQUeETR0/Ea1BMsix1wOJ1wu9yQy+QdPs/fekThwSWXY/f4OPx3Rt1P8xJZFImCiUt0SR0hJLAsdsDucLTchvYwZ1EACyNRKDs9L/GVuRYUxnC/RKJgxZFF72BZ7IDD0bLAxdNtaCIKXZyXSBQ6OLLoHSyLHXA4nRBFEYLAW05E4aJBr8LH5yZg5TjOSyQKFRxZ9A6WxQ44nU4WRaKw4kb6uBN4PVFkUSQKIW6Rq6G9gfdZO+BwOjt9nn+VEIWWxcqdmG8/ijcKiqCr50gEUahgWfQOlsUOuFz8zUUUTq5zfAIAGCaz4OPaciQ6DBInIiJvELkc1StYFokorI237sNwTWnrr9MEBz4uPYEBDqOEqYjIG5QypdQRQgLLoke82UwUDm4VVrV7zOi2Y2VxDkbbovwfiIi8RilnWfQGlsUOdLW4RRA4rE0UCtLFEsw3FnX4nAZuvFeajZkWjjASBSu1XC11hJDAstgLcpZFopDwQMR6yDs5O1YG4KXyg7jUFOG/UETkNSyL3sGy6JHnQqhkWSQKekahGee5dnTrtU9U5WBpgxYCz4UnCioquUrqCCGBZbEXFCyLREFvCT6CVrR2+/X31B7F72rlkIsdnxlPRIGHI4vewbLYgZYpi55HEFgWiYKbXHTiSmzu8ft+3XgSf6tyQCVy0jxRMFDJOLLoDTzBpRd4G5oouC3VfotU2Hv13vObixHtjsfdCVEwy2xeTkb+VvVZFRr3NsJWZoOgFKAbqEPSlUlQJ/88IiWKIirXVKLuuzq4ml3QDtAi5foUaFI1nV67YU8DKldXwl5phypBhcTLExE5IbL1+frt9Sj/qByiTUT0jGgkXZ3U+py9yo78Z/OR9VgW5FqOZvcWb0N7B0cWe4Eji0TB7QbZV316/xRLFd4prUSUW+elRCSV5iPNiJkTgwGPDED/B/sDbiD/2Xy4bT8vfKpeX42ar2qQfG0ysv6UBaVRifxn8uGyuDxe13zCjKJXihB1dhQGPjEQUWdHofDlQpjzzAAAZ5MTJW+VIPmqZPR7oB/qttWh6UBT6/tL3y1F4q8SWRT7iLehvYNlsQMtW+d4LoQsi0TB6wLVAaS4S/p8naGOBnxQUoREF1dKB7P+v+2P6BnR0KRqoM3QIvXmVDhqHLDkWwC0jCrWbKhB/KJ4GCcaoUnTIPWWVLhtbjTsbPB43eoN1TCMMCD+wnioU9SIvzAehmEG1GyoAdAycijXymGcYoRugA76YXpYS1vm0NbvqIegEGCcyG2b+kqr0EodISSwLHZAqej87jxvQxMFr2XKdV67VpqzGauKT6A/T3sJGadHC+X6lhE9R5UDzgYnDCN/PgJSppRBP1QP8wmzx+tYTljavAcADKMMre9RJ6rhtrthKbDAaXLCcsoCTboGTpMTlasrkXxtsrc/tbAUqYrs+kXUJc5Z7IBS2fnkdYWMZ0cTBaMx8gIMcx3x6jWj3TasKs7BTSkjcVBd79Vrk3+JoojyD8qhG6yDJq1lPqKzwQkAUES2/etSEamAo8bh8VrOBmeH7zl9PblejrRb0lD8WjFEu4ios6MQMSoCxW8UI2ZeDBzVDhS+UAjRJSLhkgQYJ/EHkt4wqvl18waWxQ6cHll0u92QydoPvnJkkSg4PaD+FDK39//8auDG/0qzcWfSaGzV1nv9+uQfZe+VwVpkxYCHB7R/8swNMrrz26iL90ROiGyz4MWUa4Kt2IaUa1Nw7HfHkH57OhRGBfKeyIN+iL5d+aSucWTRO3gbugMqpQJyuRxud8cjiCqWRaKgkyTUY6p7j8+uLwPwSnk2LjLxL6dgVPpeKRoPNCLz95lQxvx8d0lhbClop0cET3M2OVuf64jCqOjRe9wON8reK0PKDSmwV9ohukToh+qhTlZDnaRuXRhDPcORRe9gWeyASqWEXCaDy0NZ1Mo9r4AjosD0W+06KOHs+oV99JeqQ7iJp70EDVEUW4ri3kZk/l8mVPFtt1pRxiuhMCpgyjG1PuZ2utF8pBm6gZ5Xw2sHatu8BwBMh0we31P1aRUMowzQ9tdCdIv45SmUorPtr6n7ItX84c0bWBY7oFQoIJPJ4HJ1/KdTwzmLREFFAzsWiFv89vHuqz2KB+vkkIv8Fhvoyt4rQ/32eqTfng6ZRgZHvQOOegfc9pbv84IgIPbcWFSta9mP0VpsRcnrJZCpZTCe9fOoVfGrxShfVd7667j5cTAdMqHq8yrYSm2o+rwKpsMmxJ4b2y6DtcSKht0NSLwsEQBa9ngUgNrvatF0oAm2Mhu0A7iqtzc4sugdnADRAaVSCblc5vE2tJZlkSio3Kn9GnrR1PULvei6hpOIdabjj/EqOATPCyFIWrWbagEAp/52qs3jqTenInpGNAAgbmEc3HY3St8tbdmUO0uL/r/t32YPRHuNvc0cRd0gHdLvSEfFxxWo/KQSqgQV0u9Ihy6r7ciiKIoofasUSYuTIFO3/HAhU8mQujQVZe+VQXSISL4uGcponhrUUzJBhgglt7byBkEURU7AO0NNXQMe/vuLiNDrYIxs/xutySnHyspECZIRUW/8oLsHce4qST72Dm0C7k0w8rQXIj8zqo3YevVWqWOEBN4j6YBSIW+5Dc05i0RBb7F6h2RFEQCmWirxZlkVjDzthcivjCregvYWlsUOdLXARSEASoG3oomCwVL551JHwAh7PT4oKUICT3sh8psYTYzUEUIGy2IH1CoVVEolnE7PI4ict0gU+KYrjiLLfVLqGACAdJ72QuRXCboEqSP0WX5+PgRBwIEDByTNwbLYAUEQEBmhh8PheVI6V0QTBb57VJ9KHaGNGLcNHxYfxkhblNRRiEKeP8qiIAid/rNkyZI+XT89PR1lZWUYOXJkr95fU1OD888/HykpKVCr1UhPT8ddd92FxsbGHl2Hq6E9iIqMwKmiUo/Pa+UugAsciQJWpqwS493ZUsdoRwsX3is9iDuTRmE7T3sh8pl4XbzPP0ZZWVnrf69cuRKPPvoojh492vqYVtu3LY/kcjmSkpJ6/X6ZTIaLL74Yf/7znxEfH48TJ07gzjvvRG1tLZYvX9796/Q6QYiLNkZ2OrLI29BEge1BzaeQIzAXoykg4r/l2VjE016IfMYfI4tJSUmt/xiNRgiC0Oax5cuXIysrCyqVCkOGDMF7773X5v2CIOCVV17BggULoNVqkZmZiVWrVrU+39Ft6JycHFxwwQWIjIxEREQEZsyYgby8vA7zRUdH44477sDEiRPRr18/zJ07F8uWLcOWLT3bd5Zl0YNIgx7uTnYViuCKaKKAZRSaMVfcLnWMLv216hCWNOh42guRDyTqpN3ibvXq1bjnnnvwwAMP4NChQ7jttttw4403YvPmzW1e98gjj+Dyyy/Hjz/+iGuvvRaLFy9Gbm5uh9csKSnBzJkzodFosGnTJuzduxc33XQTnM7unU5VWlqKTz75BLNmzerR58Ky6IFOp213BvwvGRS+PzaMiHrnXs2XUItWqWN0ywO1R/BAnQIynvZC5FXJ+mRJP/6zzz6LJUuWYNmyZRg8eDDuv/9+XHbZZXj22WfbvO5Xv/oVli5disGDB+PJJ5/ExIkT8eKLL3Z4zZdeeglGoxErVqzAxIkTMXjwYNx4440YMmRIp1kWL14MnU6H1NRUREZG4vXXX+/R58LvTh7otZpOn+fIIlFgksGFS4VvpI7RIzc05OHP1W4oRU4jJ/IGmSBDol7akcXc3FxMmzatzWPTpk1rN2o4derUdr/2NLJ44MABzJgxA0plz070ee6557Bv3z6sWbMGeXl5uP/++3v0fn5n8kCna5mUKooiBKH9GGMkyyJRQLpZ8x2i3PVSx+ixRaZCRLsScX9CJCw87YWoT+I0cVDKpD8i8cz+4KlTdPW+03q7YOb0HMqhQ4ciNjYWM2bMwCOPPILk5O6NvnJk0QO9VguFQgGHo+PbzRq5mxtzEwWgG2RfSh2h16ZbKvBWWTUiedoLUZ8kGXq/gthbhg0bhq1b2x43uH37dgwbNqzNYzt37mz366FDh3Z4zdGjR2PLli2dLsDtyulTnm227v9QypFFD3RaDVRKBexOB1Sqjn86MchdqHOybxMFigWqH5HmLpY6Rp+MsNfhgxIHlqRkoEpukjoOUVDKiMiQOgIefPBBXHnllRg/fjzmzp2LdevW4ZNPPsHGjRvbvG7VqlWYOHEipk+fjvfffx+7d+/GG2+80eE177rrLrz44ou4+uqr8dBDD8FoNGLnzp2YPHlyh/MW169fj4qKCkyaNAkGgwGHDx/G//3f/2HatGno379/tz8XNh0PDHodVEol7HbP7T2Ci1yIAsqdynVSR/CKDKcJHxbnIcPBrXWIeqN/ZH+pI+CSSy7BCy+8gGeeeQYjRozAf//7X7z11ls455xz2rzu8ccfx4oVKzB69Gi88847eP/99zF8+PAOrxkbG4tNmzbBZDJh1qxZmDBhAl577TWPcxi1Wi1ee+01TJ8+HcOGDcO9996LCy+8EJ999lmPPhdBFDvZHyaMud1uPPiX52Cx2pCS2PHGnjsaIpHTbPBzMiLqyCh5EdYqfw8ZQudbmhlyLEkdgVxVvdRRiILKM7Oewfn9z5c6RpcEQcDq1atxySWXSB2lUxxZ9EAmkyEhNqbTe/pcEU0UOB7QrA2poggAOriwvOQgzrJGSR2FKKhkRmZKHSGksCx2Iik+Fhab3ePzvA1NFBgShQZMc+2ROoZPKCDitbJsLGzmLWmi7hAgICNS+jmLoYQLXDoRGxMFdHKXPoplkSggPKBdB6U7tA9rf7ryEOJihuG9SDNEIbRGUIm8KVGfCK2ib2cy+0uwzATkyGInoiIiOn0+Uu6CnNvnEElKDTsuEL+XOoZfPFibi3vrlTzthagT/SL7SR0h5PA7TieijBGQyWRwujqemygIHF0kktoyzUboxfDZYuam+hN4gqe9EHkUCCuhQw3LYieiIiOgUathtXpe5BLDskgkqcWyDVJH8LuLTYV4vsIMjVsldRSigMOy6H0si51oKYsqWDtZER2tDO15UkSB7Er1TiS4K6WOIYmZlnK8UV6LCHdwzM0i8pcBxgFSRwg5LIud0Ou0MOh1sHayIjqaI4tEkrlVsV7qCJIabavFByUliHNxv1ei04bEtD/JhPqGZbETgiAgKT6207IYw5FFIklMVRzHQNcJqWNIrp/ThFXFeUh3cmsdonhtPGK1sVLHCDksi11IS06E3e65LOrlbqi4IprI7+5Tfyp1hIAR57bho6JcDLVHSR2FSFJDY4ZKHSEksSx2ISEuBiI63wuJ8xaJ/Ku/rAoTXAekjhFQTp/2MpmnvVAYY1n0DZbFLiTExkCpkMPh8Dw3kfMWifzrQc2nkIPHbZ5JCRFvlGXjfJ72QmFqWOwwqSOEJJbFLsTHxkCv1aLZYvH4mjiOLBL5TYRgxlxxm9QxAtozlYfw60Y9BFGQOgqRXw2N5siiL7AsdiHaGIFIgwFmi9Xja+JVnuc0EpF33av5EhrR859HavH7mlzczdNeKIwYlAakRaRJHSMk8btIF2QyGTJSk9Bs9jyyGK1wQslFLkQ+J4MLlwvfSB0jaCytP4HHatxQ8LQXCgNDYoZAEDia7gssi92QlpIEp9PzvESZwFvRRP5wo2YLotx1UscIKpc2FeL5Cgs0Ik97odA2LIbzFX2FZbEbEmKjIYpipyuieSuayPeWyL6QOkJQmmUpw+tlPO2FQtuouFFSRwhZLIvdEB8bDbVKBVsn+y0mcGSRyKfOU2Yj3V0kdYygNcZWi/dLS3naC4WscQnjpI4QslgWuyEhNgZ6nbbTeYsJHFkk8qk7VeukjhD0Mh1NWFV8Emk87YVCTJI+CcmGZKljhCyWxW6IMOgRFxONpmazx9fo5G7o5dxvkcgXhstLMNJ1WOoYISHObcWq4iMYwtNeKISMi+eooi+xLHaDIAgYmtUf5k5GFgHeiibyld9q1kAGz3OGqWcMohMflBzERJ72QiFibMJYqSOENJbFbkpPSery2D8uciHyvnihATNcu6WOEXKUEPFWWTbObTZKHYWoz8Ynjpc6QkhjWeymlMR4aNQqWKw2j69JYlkk8roHtJ9DCY7a+8o/Kg9icaMeHLilYKVX6jEoapDUMUIay2I3pSTFI9KgR5Op2eNr4pQObs5N5EVq2HGh+J3UMULeH2py8Zt6NU97oaA0Om405DK51DFCGr8zdJNGrUb/9NROy6JMAJI5ukjkNbdrNsEgNkkdIyzcWn8cj9aIPO2Fgs64RC5u8TWWxR7I6pcOm6Pz22Epas+3qYmoZ66RfSV1hLByeVMB/llp5WkvFFQmJk6UOkLIY1nsgdSkBMgAuFyebzWzLBJ5x+Xq3Uh0V0gdI+zMNpfi1bI6GHjaCwUBrUKLsfFjpY4R8lgWeyA1KQEGvQ4ms+f9FmOUTmhlLj+mIgpNtyk+lzpC2Bpnq8H7paWIdemljkLUqYmJE6GUK6WOEfJYFnsgITYa0VHGTuctAhxdJOqrKYo8DHIdlzpGWBvgaMLKknyk8rQXCmBnp5wtdYSwwLLYAzKZDEOzMrssi1zkQtQ396nXQpA6BCHRZcGq4iMYxNNeKECxLPoHy2IPDeyfDrfb3enm3KkcWSTqtX6yakxy7Zc6Bv0kQnTig5JDmMDTXijAJOoSMSBqgNQxwgLLYg/1T0+BQa/r9JzoCIULETwnmqhXfqv5FHJw3m8gUcONt8uyMdfM014ocExNmSp1hLDBsthDKYnxSIiLQX1D53u/cXSRqOcMghXzxG1SxyAPnq84iCubDDzthQICb0H7D8tiD8lkMowcPBCm5s7nLWZorH5KRBQ67tF8Aa1okToGdeKR6sO4k6e9kMRkggxTkzmy6C/8094LWf3SAKHr/RYVPPqPqNsEuHGF8I3UMagbbq8/jj/WAAqRR6yRNIbFDEOUJkrqGGGDZbEX+qWnIFKvR6PJ5PE1CgFI461oom67QbMV0e5aqWNQN/2qKR/PVtmg5mkvJIHZ6bOljhBWWBZ7ISE2BsmJCahv7HzeIm9FE3XfTbL1UkegHprbzNNeSBpzM+ZKHSGssCz2giAIGDV0IMzmzstghsYKgTPBibo0T3kIGe5CqWNQL4y31eC90jLE8LQX8pN+kf0wMHqg1DHCCstiL/VPT4FMJsDp9LxFjkYmIpEbdBN16TeqdVJHoD4Y6GjEipJ8pPC0F/KDORlzpI4QdlgWe2lARhqijBGo461ooj4ZJi/BKNchqWNQHyW7LPiw5CgG2rkXI/kWb0H7H8tiL0VFRmDogP6orWvo9HX9WBaJOvVbzVrIOF0jJBjdDqwoycF4W5TUUShExWvjMTputNQxwg7LYh+MGjYYTper06P/jAoXohQOP6YiCh5xQiNmuHZLHYO8SA033irNxmye9kI+MDt9NgSBJ8f7G8tiHwzKzECkXoeGJs9b6AAcXSTy5AHt51CB83pDjQzAvyoO4oqmCJ72Ql7FW9DSYFnsg5TEeGSkJqOmrr7T1w3U8kQKojMp4cSF4ndSxyAf+lN1Dm5v0EAQORJEfRehisCk5ElSxwhLLIt9IAgCxo0cBqvV1umt6GilEzG8FU3Uxu2abxAhNkodg3zszrpjeLhWgJynvVAfzcuYB6VMKXWMsMSy2EeDB2RAq9HAbOn8VvNAndlPiYiCw69lX0kdgfzkqsZ8PFNlg1rkX/TUexcOuFDqCGGLZbGP+qelICk+tstb0VlaCzh5h6jFpaofkOQulzoG+dH85lK8Ut4AvVsjdRQKQkn6JExK4i1oqbAs9pFCocC4kUPRZGru9HV6uRtJ3KCbCABwu/IzqSOQBCZZq/FuaTmiedoL9dDCzIVcBS0hlkUvGJLVHwqFAjZ752WQC12IgImKkxjsOiZ1DJLIYEcjVpYUINkZIXUUCiKLBiySOkJYY1n0gkGZGUiIjUF1bX2nr8vUWrj5MIW9B9RrwfGB8JbsMmNVyTFkObgXI3VtSPQQngUtMZZFL9BqNJg0ZjjqGzo/+k8tE5HOPRcpjKXLajDJtV/qGBQAjG4HVhbnYAxPe6EucGGL9FgWvWTUsMFQKRWwWDsvg1m8FU1h7LeaT6GAU+oYFCDUcOPd0mycY+EII3VMJsiwcMBCqWOEPZZFLxmc2Q+pSQmorKnr9HUZGivUgttPqYgChw5WnCtukzoGBRgZgBfLD+IyE+cwUnuTkyYjQZcgdYywx7LoJUqlAlPGjUJTU3OnG3QrBGAQ91ykMHSP9itoRf7ep449XpWDW3naC53h0oGXSh2BwLLoVSOHDoRep4GpufO/EIeyLFKYEeDGr4SvpY5BAe43tcfwUK2Mp70QACBGE4P5/eZLHYPAsuhVmemp6JeWgqoubkVHKZ1IUtn8lIpIetdptiHGXSt1DAoCixtP4W9Vdqh42kvYu3jgxVDK+fsgELAsepFMJsOUcaNgtlg6vRUNcHSRwsvNsvVSR6Agcn5zCV6uaORpL2FMgIBfDfqV1DHoJyyLXjZicBaMEQbUN3a+jU6m1gK1zOWnVETSmaM8jH7uAqljUJCZYqnCO6UViHLppI5CEjgr+SykR6ZLHYN+wrLoZalJCRg0oH+Xt6LlAjCY2+hQGPiN6lOpI1CQGuJowMqSQiTxtJew86shHFUMJCyLXiYIAiaPHQGH0wmnq/ORw6H6zs+TJgp2g+VlGOM6JHUMCmIpLjM+KjmOATztJWzEa+MxO3221DHoF1gWfWDs8CFIjItBVXXnE/qNCheSudCFQtiD6jWQgfuKUt8Y3XasLM7BaJ72EhYuHXQpFDKF1DHoF1gWfSAywoCzxo1CTV1DlwtdhnF0kUJUjGDCLPcuqWNQiNDAjfdKszHDEiV1FPIhmSDDFYOukDoGnYFl0UcmjhnRrT0X+2us0HOhC4WgB7SfQwW71DEohMgAvFyejUt42kvImpM+B8mGZKlj0BlYFn1kQEYaBg/oh7LK6k5fJxOAEQaTn1IR+YcSTlwkbpY6BoWoJ6tycHODlqe9hKAbRtwgdQTqAMuij8hkMkybOA5OpxNOp7PT1w7VmaHkedEUQm7VbEKE2Ch1DAph99YexYN1cshF/jUWKsbEj8HYhLFSx6AO8E+ZD40ZPhiJ8bGo7GKhi0omcpNuCinXyr6SOgKFgesaTuKvVU6e9hIiOKoYuFgWfSjCoMfU8aNRW9/1QpeRBhNk6Pw1RMHgItVeJLvLpI5BYWJhczFeqmiCzq2WOgr1QZohDXMz5kodgzxgWfSxiaNHwKDXocnU+apnvdyNTG7STSFgmfIzqSNQmDnLUom3y6oQ5eZpL8HquuHXQSawkgQq/p/xscyMVAzJ6t/lQhcAGM2FLhTkJihOYYjrqNQxKAwNs9fjg5IiJLq4UjrYGNVGXDLwEqljUCdYFn1MEATMnDIBbrcbNnvn24jEKp1I4SbdFMTuV68F16eSVNKczVhVfAL9edpLULly8JXQKTkqHMhYFv1g7PAhyExPRWl5ZZevHcXRRQpSKbJaTHHtkzoGhblotw0fFh/GKJ72EhRUMhUWD10sdQzqAsuiH6jVKsw+exKaLdYuz4tO19gQo3D4KRmR9zyo+RQKdL5NFJE/aOHC/0qzMZ2nvQS8SwddinhdvNQxqAssi34yacwIJCfEobwbcxfHRTT5IRGR9+hgxXniVqljELWSAXilPBsX8bSXgKWSqbB01FKpY1A3sCz6SWSEAeecNRH1DY1wuzvfgLu/xopoji5SEPmNdgN0IvcKpcDzl6oc3Nig42kvAejSQZciSZ8kdQzqBpZFPzprwmjERBlRXVvf6esEgaOLFDwEuHGlsFHqGEQe3V97BL+tU/C0lwDCUcXgwj85fpQYF4upE8agsrq2y026MzVWRHF0kYLAr9U7EOvuenoFkZSub8jDn6tcUPK0l4DAUcXgwrLoZ9MmjoVBr0NDY+erngUBGM/RRQoCNyvWSx2BqFsubC7CvytM0PK0F0kpZUqOKgYZlkU/y8xIxdgRQ1BWWdX1azVWroymgHaO8ggyXaekjkHUbWdbKvBWWRWMPO1FMpcNuoyjikGGZdHPWjbpHg+FQgFTc+cLAgQBmMDRRQpgd6vWSh2BqMdG/HTaS4LLIHWUsMNRxeDEsiiBkUMGYtTQQSguq+jytf20VsQrOz/5hUgKg+TlGOs6KHUMol5JdzZjVXEeT3vxsysGX8FRxSDEsigBmUyGedOnQC6TwWTuersRji5SIHpQvRYydL4NFFEgi/nptJcR9iipo4QFg9KA28fcLnUM6gWWRYmMGvrT6GJp16OLaRobz4ymgBIjmDBL3Cl1DKI+08KF/5UcxFRrlNRRQt6NI29EjCZG6hjUCyyLEpHL5Zg34yzIujm6OMXYAAGdb7dD5C/3adZDLfIHGAoNCoh4tSwbF5oipY4SshJ0Cbh++PVSx6BeYlmU0KihAzG6m6OLsUonBul4QgZJTwEXLsZmqWMQed1TVYewhKe9+MRdY++CRqGROgb1EsuihORyOc6dNRVyuRxNpuYuXz8xogkKgXPESFq3aDYjUmyQOgaRTzxQewT31ykg42kvXjMwaiAuHnix1DGoD/inQWIjhwzE2OFDurUyWid3Y7Sh8828iXztOtmXUkcg8qklDXl4stoNpaiQOkpIuH/C/ZAJrBvBjP/3JCaTyXDurKlQqZRobOq6CI7WN0Mnc/khGVF7F6r2I8VdKnUMIp+7yFSIf1U087SXPpqSPAUz0mZIHYP6iGUxAAwfNAATRg1HUVlFl2dGK2QiJkY2+ikZUVvLlOukjkDkN9MtFXijvBqRbq3UUYKSAAH3T7hf6hjkBSyLAUAQBJw362wYdFrU1nc9F2yQ1oJYbtRNfjZWXoChrqNSxyDyq1G2OnxQUoJ4nvbSY5cNugzDY4dLHYO8gGUxQAzsn44Zk8ejtLwKbnfni1gEAZjC0UXyswfUayDj9k0UhjKcJnxYnIcMB7fW6S6j2oh7x98rdQzyEpbFACEIAs6ddTYS42NQXlnd5etT1HZkaix+SEYEJMvqMNW9V+oYRJKJc9uwqjgXw3jaS7fcPe5uRGmipI5BXsKyGECS4mNx7sypqK1vhMPp7PL1ZxkboORWOuQHD2rWQYGuf08ShTIdXFhechBTeNpLp0bGjsQVg6+QOgZ5EctigJl11kQMyEhFUWl5l6/Vy908N5p8Tgcbzhe3SB2DKCAoIOL1smwsbOYt6Y7IBBkePuthbpUTYvh/M8BEGPRYOGcGbDY7LFZrl68frm9GHBe7kA/dqd0Andj1pvFE4eTpykO4rlHP017OcNmgyzAybqTUMcjLWBYD0JRxozByyEAUFJd1+VqZAEzjudHkQ1cJG6WOQBSQ/q8mF/fWK3nay0+i1FFc1BKi+Ds8ACmVClwwdwaUCgUaurFRd7zKgWF6jvyQ912j3o44d5XUMYgC1k31J/B4DU97AYB7xt8Do9oodQzyAZbFADVq6CBMGTcSxaXlXW7UDbScG82TXcjblsrXSx2BKOBd0lSI5yvM0LhVUkeRzPiE8bh80OVSxyAfYVkMUIIg4IK5MxEdZUR5VU2Xr1fJREwxdr2hN1F3zVAexQD3SaljEAWFmZZyvFFei4gwPO1FI9fgiWlPQBA4fzNUsSwGsPSUJCw4Zxpqauthtzu6fH2W1oo0ddeLYoi6417VWqkjEAWV0bZavF9airgwO+1l2dhl6BfZT+oY5EMsiwFu7vQpGDooE/lFJd16/TTuvUheMFBWgbGubKljEAWdTEcTVhWfRLozPLbWGRU3CtcPv17qGORjLIsBTqfV4NLz5kCukKOuoesj/iIULh4FSH32W81ayMEfOoh6I85txYfFRzA0xE97UcqUeOLsJyCXyaWOQj7GshgERg8bhBmTx6O4rKLLc6MBYKjezNvR1GtRQjNmizukjkEU1AyiE8tLDmJyCJ/2cuvoWzEweqDUMcgPWBaDgCAIWDRvFlITE1BUVtGt98yIqoeat6OpF+7VrIdatEkdgyjoKSHijbJsnN8cetvJDI0ZiqWjlkodg/yEZTFIxMdGY9H8WTCbLTBbuh411MvdOIuro6mHFHDhUmyWOgZRSHmm8iB+3ahHqJydoBAUeOLsJ6CQcW/JcMGyGESmTxqHcSOGIr+opFt7Lw7SWdBfY/FDMgoVN2u+hVGslzpGwHpqiw2TXjMh4qlGJDzThEtWmHG0uu3+pqIo4rFvrUj5RxO0f2nEOW83I6ey6z1QPz7swPCXTFD/uRHDXzJhdW7bHRDez3Yg/bkmxDzdiAc3tP2BMb/ejcEvmtBoC5E2EoJ+X5OLe+pVIXHayy2jb8Gw2GFSxyA/Cv7ftWFEqVTgkvNnI8KgR1VNbbfeM83YAA0366Zuul72pdQRAtp3BU7cOUmFnTfr8fV1OjjdwLn/M6PZ/nNJ+/s2O/65w45/L9Rgzy16JBkEzH/PjKZOityOIieu+siC60Yr8ePtelw3WokrP7JgV7ETAFBtdmPpOguena/BV9fq8c6PDnx+7OcyecfnFvxtnhqRau5zF8iW1p/AYzVuKIL4tJcx8WNw2+jbpI5BfsayGGQG9s/AuTOnorK6Dja7vcvXa+VuTOftaOqGhaofkeru3hZN4erLa/VYMlaFEQlyjEmS462LNShsELG3rOUHMlEU8fwuOx6eocZlw5QYmSDHO5doYXaIWH7Q816pz++yY36WHA/NUGNoXMu/52bK8fyulj/jJ+tEGNUCrhqpxKRUOWZnynG4qmVO8vKDDqjkAi4bpvT9F4D67NKmQjxXYYFGDL7TXgxKA/42429c/RyGWBaD0ILZ0zFy6ECcLCju1u3o/lorBmrNfkhGwWyZcp3UEYJOw0/rgGK0LSN6p+pFlJtEnJv188iRWiFgVn8Fthd7HuHfUeTCuQPajjadl6XA9qKW9wyKkcHsELG/zIVai4g9JS6MTpSj1iLi0c1W/HuBxsufGfnSOZYyvFZWF3Snvfxhyh+QFpEmdQySAMtiENJpNbhq0XmI0OtQUd31UYAAcLaxAZFyp4+TUbAaLS/EcFeu1DGCiiiKuP8rK6ZnyDEyoWWkpdzUMtqXaGh7OzhRL7Q+15Fyk4hEQ9tvx4kGGcpNLT8MRmsFvHOJFtevsWDyayZcP0aJ8wYq8NsNVvxmsgqn6t0Y918TRr5swkeHuz7tiaQ31laD/5WWIjZITntZkLkAi7IWSR2DJMKyGKQGZWbggrkzUVNbD6ut621OVDIRc2JqIQ+V5XjkVb9Vr4WMvzd65K71VmRXuPDB5e1Hh86cOSiK7R/r6XsuHabEwTsMOHF3BB47R4Nv8504WOnCLRNUuPojC54/T4OPr9Ti5k8tqGzmtlnBYICjCR+WnERagJ/2kmJIwSNnPSJ1DJIQy2IQmz/zLIwdMaTbt6PjlE5M5ukudIYkoR5T3XukjhFUfrPegk+PObH5Bj3SIn/+Npr00+jg6RHB0yrN7UcOfynJ0H7ksbLZ3W6E8jSbU8Syz63474VanKh1w+kGZvVXYEicHINjZdjVyS1vCiwJLitWFR/B4AA97UUuyPHU9KcQoYqQOgpJiGUxiGnUalx54XmIioxAWUVVt94zwtCMftxOh37ht9p1UIJTFLpDFEXctd6CT444sel6HTKj234LzYwSkGQQ8PXJn7+edpeI7/KdODvN86KAqelyfH2ybcHbcNKJs9M7fs+T39uwYKAC45PlcLkBp/vncupwAS4OEgcVg+jEipKDmBiAp70sHbUU4xPHSx2DJMayGOQyM1KxaP45qG9s6tZm3QAwM6oeBs5fJAAa2LFA3CJ1jKBx53or/pftwPLLtIhQt4wGlpvcsDha2pkgCLh3igp/3WLD6lwHDlW6sGSNBTqlgGtG/bxa+frVFjy08ec/r/dMUWFDnhNPb7XhSLULT2+1YeNJF+6d0n7FbE6lCytznHhithoAMDROBpkg4I19dnx+zIEj1W5MSuFq1WCjhIi3yrIx3xw4p71MSJyA28fcLnUMCgDBu9kTtZo7bTIOH8/Dnh9zMGJwFgSh89lRapmIOdF1+Kw6Du4uZ1JRKLtT+zX0oknqGEHjlR9aFo+c807b3QXeuliDJWNbit3/TVPB4hSxbL0VdRYRU9Lk2HCdDhG/2AOxsMENmfDzz+pnpyuw4got/rjJhkc225AVI8PKK7SYktb2W7Qoirj1MyueO08NvarlelqlgLcv0eDO9VbYnMC/F2qQGslxgGD1z4qD+EvsMKyIaO56oqsPxWni8OysZ3lKCwEABLE7k90o4BWVluOZ/7wNq82OjNTkbr0n26TH7sbA+SmW/G+P7l7EuyuljkFEZ/hv1CC8HOWAW/D/YiW5IMeb573J28/Uij9+hoj0lCRcvnAeLFYbGhqbuvWeUfpmpKu7d+uaQs/V6h0sikQB6rb643ikBlCI/p9ScN+E+1gUqQ2WxRAyY/J4zD57EopKy+FwdD0nURCAWdF1iOD8xbC0VLFe6ghE1IkrmvLxj0qbX097mZcxDzeMuMFvH4+CA8tiCJHJZLhi4TwMGzQAJ/ILu7WdjkYmYn5MLRQS3Oog6UxTHMNAV57UMYioC3PMpXi1rA4GP5z2kmHIwJ+n/9nnH4eCD8tiiImMMODXly6EMTICJeXdu8UYo3RiVlQ9wE2Zw8a9qk+ljkBE3TSu9bQXvc8+hlqmxr/m/gt6pe8+BgUvlsUQNLB/Bi5fMBemZjMaTc3dek+m1opxBq6KDQeZskqMd/8odQwi6oEsRxNWluQj1UenvTw+7XFkRWX55NoU/FgWQ9Q5Uydi5pQJKCwuhcPZvTmJ4yOakMENu0Peg5pPIQdP+CAKNokuC1YVH8Egu3d3sVgybAkuGHCBV69JoYVlMUTJ5XJcuehcDMnKxIlT3Zu/KAjAOVH1iFI4/JCQpGAUmjFX3C51DCLqpQjRiQ9KcjDeS6e9zEiagfsn3e+Va1HoYlkMYVGREbjmkgWINOhR2s3jAFU/LXhRc8FLSLpX8yXUIrdLIgpmarjxTlk25vbxtJdMfSb+MfcfXR7kQMSyGOKGZPXHZQvnwWRqRn1D9/ZfNCpcmB1dB4ELXkKKDC5cKnwjdQwi8pLnKw7iyiZDr9YmRsoj8dqC16BV+H6VNQU/lsUwMHfaZMyfORXFZRWw2mzdek+axoYpkY0+Tkb+dLPme0S566WOQURe9Ej1YSxr0EAmdv+vcwUUeGX+K0jUJ/owGYUSlsUwIJPJcMUF8zFx9HCcOFUIl6t7ixtGGpoxUs8V0qHiBtkXUkcgIh+4o+4Y/lgDyLtz2osIPDblMYxOHO37YBQyWBbDhE6rwfW/WoQBGWk43s0FLwAwJbIRA7hCOuidp8xGmrtY6hhE5CO/asrHM1U2qEVlp6+7NutaXDz0Yj+lolDBshhGEmJjcP0VixBp0KOotKJb7zl9JGCSqnu3rykw3aVaJ3UEIvKx+c2l+G9ZA/RuTYfPT4ueht/N+J2fU1EoYFkMM0MHZuLqi8+HzW5DTV19t94jF4D5MbXcUidIjZIXYYTrsNQxiMgPJtiq8b/ScsSccdrLEM0QvHjBixKlomDHshiGZkwej4Wzp6OsshrNlu7dYlbLRJwfUwudjJs5B5sHNGsh48p2orAx0NGIFSX5SHIYAADJ8mS8fcnbUMo7v0VN5AnLYhgSBAGXnDcHZ08Yg5MFxd0+4cWgcOG82BoouQdj0EgUGjDNtUfqGETkZ8kuC1YVH8MkRwrevvBtGNQGqSNREGNZDFNqtQrXXnYBhg0agGN5BXC7u1cAY5VOzI2u40hVkHhAuw5KcPoAUTiSCRr8Y/4zSIlKkToKBTmWxTAWE2XE0qsvRXpKIo6dLOj2Cuk0jQ2zuGl3wFPDjgvE76WOQUQSaIYaDZe+jeh+3CKH+o5lMcylJSfi5qsvRUxUJE4VlnT7fVlaK2ZE1aNXRweQXyzTbIRe5D6ZROHGKipQOOefSB89X+ooFCJYFglDsvrjussXQSGXo7ise1vqAMBgnQXTjA0+TEZ9sVi2QeoIRORnDlGGQ+N/j2Ezr5U6CoUQlkUCAEwaMwJXXXQerFYbqmrquv2+YXozpkSyMAaaK9W7kOCulDoGEfmRQ5Rh19C7MX7RA1JHoRDDskit5kybjIvOPQfVtXVoaGzq9vtGGZoxIYLnSAeSWxSfSx2BiPzIIcrw/YClOOtXj0Am41/t5F38HUWtBEHARfNnYd70s1BYWt7tPRgBYFyECWMM3S+Y5DtnKU5gkOuE1DGIyE8cbuBT40WYdvWTUCgUUsehEMSySG0oFAosvuR8TJ80Difzi2G12bv93kmRTRip54IKqd2nXit1BCLyE5coYIvhAsxZ8g9o1B0f80fUVyyL1I5Grcb1VyzC+FHDcPxUAez27u/Td5axEcN1zT5MR53pL6vCRNcBqWMQkR+4RQHbIy7AxFteRHRMnNRxKISxLFKHjBEG3HLNZRgzbDCOncyHw9G9U14A4OyoBoziCKMkHtR8Cjl4JCNRqHOLArZq5mHE9c8iMjpW6jgU4lgWyaOYKCNu/fUVGD44C0fz8rt9LCAATDE2YhznMPpVhGDGXHGb1DGIyMdcooDvlOdg6PX/QExCstRxKAywLFKn4mKicNu1V2DowP44lpcPp6v7o1YTIpswmdvq+M29mi+hEa1SxyAiH3KIMnyjnIeh1/8DCan9pI5DYYJlkbqUGBeLW399BbL6p+PoiXy4XN07RxoARhuacbaxHjzpxbdkcOFy4RupYxCRD9lEOb6Sn4eR1/0NyRlZUsehMMKySN2SkhiP26/9FTLTU3A0Lx9ud/cL43C9GTOj6nmWtA/dqNmCKHf3N1MnouBicSuxXr4A4657Ain9Bkodh8IMyyJ1W1pyIm799RVIS07ocWEcrLNgdnQdZCyMPrFE9oXUEYjIR0xuNdYpFmLStY8iNXOw1HEoDLEsUo/0T0/BbddegeSEOBzLK+hRYRygtWJuTC3kLIxeNV95COnuIqljEJEP1Lu0WCdfiKm/fghpA4ZIHYfCFMsi9VhWv3Tccf2VSE1KwJETp+DqwaKXfhobFsTWQC3j9i7ecpfqU6kjEJEPVLsM+FR+AaZd8yDSs4ZJHYfCGMsi9UpWv3Qsu+FK9E9PwZETp+DswbY6SWo7LoqrRoS8+++hjg2Xl2CUK0fqGETkZUWOaKyTX4BZ19yHjEEjpI5DYY5lkXqtX1oKll1/FQZlZuDIiZ5t3G1UuHBRXDXild0/TpDa+61mDeeBEoWYHFsyNmguwvzr7ke/wSOljkMEQRRF/k1DfVJeVYNX3/8IOcfyMGRAf6hUym6/1+kWsKk+CoVWrQ8ThqZ4oQHb1XdDie4fx0hEgW2beQCOxszHwl/fjqT0AVLHIQLAkUXygqT4WNxx3ZUYM2wwjp7Mh9XW/dFChUzE/Og6nifdCw9oP2dRJAoRbsjwedNw5CVdiIuW3M2iSAGFI4vkNXUNjXht+SfYd/AwsvqnQ6vR9Oj92SY9djdGAhB8EzCEqGHHXu1vYBB5pCJRsLNDhY/qRwJZs7Fg8W2IjkuUOhJRGxxZJK+JNkbitmuvwORxo3AivwimZnOP3j/a0IzZ0XXcWqcbbtdsYlEkCgEmwYC3a8dDOfQ8LLruLhZFCkgcWSSvazZb8L/Vn+PbHT8gOSEOMVHGHr2/2q7E13XRaHYpfJQw+O3S3YdEd4XUMYioDyqFBKyoHoK0sXNw/lVLoTNESh2JqEMcWSSv0+u0uPFXF+Oi+bNQVV2L8srqHr0/TuXAJXHVSFLZfJQwuF2u3s2iSBTkcjEIb1ePwcCpF2Lh4ttYFCmgcWSRfMbtduOrb7fjo/VfQ5DJ0C81GYLQ/fmIbhHY2WDEYbPehymDzwb9nzDYdVzqGETUCy7I8L1rHPaYkjFl7iJMO+9yKJTd30GCSAosi+RToihi575svPfJ5zA1N2Ng/wzIZD0b0D7arMP2BiNcXPiCSYo8fKh4hF8JoiBkEXRY2zwe5UIiZl54NcZOm9fj74dEUmBZJL/IOZaHt1auQWlFFQYP6AeFomfzESvtSmysjYHZLfdRwuDwgf6fmOr6QeoYRNRD1fJErKgZDllUKuZddgMGj54sdSSibmNZJL8pKC7FGytW4+jJAgzK7AeNWtWj95tdMmysjUGlo2fvCxX9ZNXYpLoPcvBcbaJgclw5HJ+UpSAuLQvnX7UUqZmDpY5E1CMsi+RXVTV1eOvDtdh78DD6pSYjMsLQo/e7RGB3YyRymnv2vlDwou5NLHJvlDoGEXWTEwrsUp6Nb0vVyBw6GudduRSxiSlSxyLqMZZF8jtTsxkfrvsKm7bvhjEyEskJcT2+RoFFg+/ro2ATw2O+j0GwYo/mTmhFi9RRiKgbGuQx+NI1DSdrbBg+YTrmXb4E+oiebSNGFChYFkkSLpcLG77fiU++2AiH04kBGWk9nuhtcsmwuS4aFXa1j1IGjoe1q3GLuErqGETUDXmqEVjfkAWbE5g8+0JMPfdSKFXhOX2GQgPLIklqf84RvP/J5ygpr8SgzH5QqXq2hYRbBPY2RSDbZIAYomuEBbixT3c3ot21Ukchok44oMQu7RxsLZEhIjoGsxddg2ETpvVoyzCiQMQjMkhS40YMRXxMNN5ZtQ4HjxxDv7QURBi6v6+iTAAmRTYhRW3Dt3XRsITgaukbNFtZFIkCXK08HpuUc3GssAZpAwZh3hU3IqXfQKljEXkFRxYpIDSZmvHB2i/x3c4fEBNlRGJ8bI+vYXbJ8F19FEpsGh8klM53uofQz10gdQwi8uCoZiy+s41EdVUlhk+YhrmXXo+IqBipYxF5DcsiBQyn04kvNm/D2g2b4XS5MSAjtcfzGEURONisx97GyJDYxHuuMgdvyP8idQwi6oBF0GOXfj72V8nhcjo4P5FCFssiBRRRFLH3YC5WrP0CxWWVGNAvDTptz0cK6xwKfFcfheog35Nxjf4pjHUdlDoGEZ0hXzUEO1QzkF9QBGNsPGZduBjDOT+RQhTLIgWksspqfLBmPfb8mIP42BgkxPX8lo5bBH40GbC/KQLuIBxlHCovxXrl/0EGt9RRiOgnVkGLPfq5OGyJQ015CQYMG4vZl1yLxLT+Ukcj8hmWRQpYNpsd6zdvxfpNW2CzO5DVLw1yec8XsNQ4FPi+Lho1zp6ttJbaG7qXMde9VeoYRPSTYmUWdurmoqC0EqLbjQkzzsPZ510GtVYndTQin2JZpIAmiiJ+PHwMKz79EvlFJeifkQqDruffmN0isL8pAgeCZIudOKER29V3QwW71FGIwp5dUGOvbjaOYgBKC04gLjEVMy+8CkPGnsXbzhQWWBYpKFTV1GHluq+wfe8BREdGIikhrlffpKvtSnxXH4W6AB9lfEr3ARa710kdgyjslSozsUs/H+UNVtRXlWPgqImYc8m1iE1MlToakd+wLFLQcDic+HrLDny64VuYzBZk9U+HUtHzrUJdP81l/LEpIiBXTCvhxD7tXYgQG6WOQhS2zIIee/Wzka/IQkXRSchkCkyafQGmzF0ElTq0tuci6grLIgWdw8dPYsXaL3D0ZAFSEhMQG92781YbnHJsbzAG3L6Mv9F8hQfwjtQxiMKSGwJOqEfjgG4G6k0WVJUWIjkjCzMvuBJZI8bztjOFJZZFCkqNTSZ8+vV32LRtN5wuFwZkpELRi1FGAMizaLCrwQhzgJz+slN3P5Lc5VLHIAo7dfJ47NbPR5UsARXFpyC6RYyaMgvTzr8cBmO01PGIJMOySEFLFEUcOHwUH3++ESfyC5GanIhoY2SvrmV3C9jbFIHDzXpJF8BcovoBz8v+KdnHJwpHDihxUDcVRzUT0NxsQkVxPhJT+2H6giu4iIUILIsUAuobm/Dphm+xefseiBDRPz0Vil5ssQO0LIDZ1mBElUSbeX+lewxD3Mck+dhE4ahIORB79bNhEgyoLCmA027DiEkzMH3BFTDGxEsdjyggsCxSSDh98ssn6zfiZGEx0pKTEGWM6OW1gCNmHX5ojIRN7Nlxg30xUXESqxR/DMAlN0Shp04eh3362ahQZsDSbEJF0SnEJaXi7PMvx/AJ03p81ChRKGNZpJBS19CINV9uxrc7f4AgCOifntLrUUabW8CBpgjkNOv9cgLMcv3zONu12+cfhyicWQUtsnXTkKceBadbRGVJPlwOB4aOPQszLrgS0fFJUkckCjgsixRyRFHEngOHsOarzThZWIyE+FjEx0T3et5Ro1OOPY2ROGXVejnpz9JkNfhWdR8UcPrsYxCFMxfkOKoZh0Pas+CUqdFYV4OaimIkpmXi7PmXYMi4qb06IYooHLAsUshqaDLhq2+34Zutu9DUbEH/9BTotL3fJqfCrsSuBiMqfTCf8QXdW7jY/bXXr0tELfMS9+tnwSSPgsNuQ3nhSajUGow+azamzF3Elc5EXWBZpJB3/FQhVn+5CT8ePgqdVoP0lKQ+zUc6adFgT2Mkmly926rnTDpYsVd7F7Si2SvXI6IW1YpkHNBNR6UyA6IooraiFKbGOvQbPBLTzrsMGYNGcKUzUTewLFJYcDic2LpnPz7b+B1Kyiv7tM0O0HIKzOFmPQ40RfR5EcxD2rW4TVzZp2sQ0c/q5PHI1k1DiSoLAGBpbkJlcT4io+MwafYFGDttHk9hIeoBlkUKK1U1dfhs4/fYsnsfHA4H+mekQq3q/W1lu1tATrMeh0yGXpVGAW7s1d2NGHdtrzMQUYtGWTQO6s5GgWoIIAhw2O2oLMkHRBFDxkzB1PMuRXxyutQxiYIOyyKFHVEUcfDIcaz5ajNyj59EhEGP1KREyOW9HyG0uwUc+qk02ntQGq/XbMETeKXXH5eIgGZZBA5qp+KUegREQQa3242a8mKYTY1IzRyCs+YtQtaICVzAQtRLLIsUtixWK7btOYAvNm9DSXkF4mNjkBAX06c5TD0tjd/q/4D+rvxefzyicGYW9DisnYwTmjFwC3KIoojGumrUVpYhLjEFE8+5ACMnz4Ra47udDIjCAcsihb26hkZ8s3UXNm3bjdr6RqQlJ/Z6Q+/TbG4Bh0wG5DTrPZbGc5SH8bb8z336OEThqEkWhcPaSTilHg630LLQzNLchMqSAugMkRg95RxMmHk+ImPiJE5KFBpYFol+UlxWgfWbtmLX/mxYbXZkpCVDr+3biERraTRpYUfb1dOr9X/DOFd2n65PFE7q5PE4rJ2MQtVgiELLD2EOuw2VJQWACAwaNQGT5y5CSr+BEiclCi0si0S/IIoick+cwvpvtiA79xhkchkyUpOhUir7dN0DO75BvSIWNRlzUOfSYLC8DF8qH4QMbi8lJwpdlYpU5GinoEyV2fqY02FHVWkRHHYr0jKHYPLcCzFw5EQe00fkA97ZKI4oRAiCgOGDBmBwZgZ+yD6M9Zu24PipQui0GqQlJUKp7PkfGbOpEWL1SZw1wIGhiTk4ZInGdMc2yJwsikSeiABKlQOQo52MamVq6+MupxPV5UWwmpuRnJGFibMWYMjYKVCq1NKFJQpxHFkk6oTZYsXOfdn4estO5BeVQKfTIi05EUpF90vj0b1bIZYfwtx550Gt+Xlvt3hHMYZY9yHNfgIy8I8hEQDYBRVOqUfgmGYsmuQxrY+7XC7UVJTA3NSAxLT+GD/jPAwffzbUWp2EaYnCA8siUTc0my2tpbGwpAw6rRapyQldlkaH3YaDG5ZjQFI0xk+a0uFr9K4GDLHuwwBbDlSizRfxiQJegzwGxzRjcUo9Ak7h571P3W43aitLYaqvRVxSGsbPPA/DJ0yHztC3RWhE1H0si0Q98MvSWFBcCr1O12lpPHV4H8x5uzB79lwYIjr/y00uOtDPdhQDbdmIc5b5Ij5RQHFDQKkyE8c041Cu7Af8Ytsql8uJusoymBrqEB2fhHHT52Pk5FkwREZJF5goTLEsEvVCs9mCHXt/xMatu5BfXAqDXofUpLal0e124cevViA1Uo6zps/q0fWjnJUYaM1Gpj0XStHu7fhEkrIKWpxUj8BxzVg0y41tnnM6HKipKIG1uQkxiSkYM3UOho+fxm1wiCTEskjUB6ZmM3buy8bGrbtQWFIGtUqF1OQEaNRqlJw6itrsTTj77KmIT0js1fXlogP9bbkYaM1GrKvCy+mJ/McNGUqVmTipGYES5QCIQtvTVOw2K2rKi+GwWxGf0g9jz56HoePOgj7C6OGKROQvLItEXtBstuCH7Bxs3vED8k4Vtjx2ai8SFGbMnDPfKx8j2lmBAbYc9LMdgUa0eOWaRL5WL4/FSfVI5KuHwSrTt3veZjGjuqwIbrcLyf0GYezZczFkzGQuXCEKICyLRF5ktzvwY+4xbN66Ewc3vIe0KC36ZQ1CRFRsn44R/CVBdCPZkY/+tlyk2U9AAadXrkvkLXZBjQLVUJxUj0CNMrnd86IowtRYh/qqCshkcqRnDcXYaXORNWIClCpVB1ckIimxLBL5gNvtxoncg8jL3onjB3+AqaEOEdGxiIpLglwu7/oC3aQQ7Uizn0B/Wy6SHAXcgock44ASJaoBKFQNQakqs/UYvl9yuZyor6pAU0MNdAYjBgwbixGTpqP/4FGQ92A7KiLyL5ZFIh+rLi/Gkf07kbPne9RWlUOhVCEmIQVavcGrH0fjbkY/21Fk2I8i1lnG4kg+54ASpaoBKFANQZmqP1xCxycd2Sxm1FSUwGG3IiY+GcMnzsCQMZMRn5LhtRF3IvIdlkUiP2luakBezn7k/LAFpfnHYbdaERkTC2NMAmReHG0EWopjqj0P6fYTSHQUQg6XV69P4aulIGa2jiB6KoiiKKKpvgb11RWQK5RI6TcQIyfPwqBRE6AzRPo5NRH1BcsikZ+5XC6U5h/DsR/34Gj2LjRUV0GpViMmIQUaXfsFAH2lEO1IsZ9Emv0EUhynoOJWPNRDzTIDSpUDUKrKRLmyn8eCCLSMItZVl8NqNsEQGY2sEeMxfMI0ZAwczlvNREGKZZFIQqaGOpzI2YecPVtQVpgHh92GyOhYRMbEQy73/l+sMtGFREchUu15SHIU/n97dxYbVd3wcfx3Zt+HLlPaQsFWEAqGhweXR19EXKKJ8UISEyIJiULickGi0XihURIvvPIKTSRqjBrFYEjgwnBn8EKDxEdeIVKx+hYp0IXu05npLGd7L1qKFc7zVNoC0u8nOZkzh7MNF+XLv+ecUcoZnvVj4O/PkaGBQIO6Qy3qDrZoJJD5j+vbtqXsYL9yI4MKBkPKLFqqVbetV0vrWtUsbLxKZw1grhCLwHXAtm11nfpF7cf/rd9O/KDRoX75DJ9SNRkl0tXy+Xxzcty4nVW92akGs1MLzTMKu6U5OQ6ufyUjqp7gTeoONasneJMqvuh/XN91XRVyIxoZ6JNjm0pX1+mWNXdq2a3rtLhlJaOIwA2EWASuM4VcVp2/ntBvP/2gzt/alM8OKxSOKF1Tp1giNXc3BLiuaqxe1Zudqjc7VWt1yy9nbo6Fa65kRNUXXKy+QJPOB5uU9ddM+bq9y3FdV6WxgrJD/SoVcool01rcskKt6/5HzSvXcC0icIMiFoHr2PDAeXX+ekK//HhEPWc6VMyPKppIKl1Tp3AkNqd3kvpdUzVWrzJmlzJWt2qtboXc8pwdD3NrPA6bdH4iEKcTh9LFQBwd7lexkFMkGldtQ5NW/ONfamn9h2obmrijGbjBEYvA34Druurr6tTp9p/0y4/faaD3nMqloiLRuFLVtYrGk3P/D7brKm0PKmN1qdbqVsbsVtIZmdtj4oo4MpT112owUK+BQIMGAo0aDdRMe3vXdVUujo8gFgs5hSNxZRqbdMuaO7Rk+WotXNw8q88LBXB9IxaBvxnbstR79pTOnWrXbz/9oP6esyoWcgqGwkpV1SqeWjBn1zj+WcQpqNrqVZXVryq7T1VWnxJOVowzXT2upJyvSkOBOg0F6jUUWKihwEJZxl/7JhTXdTWWyyqXHVJpLK9wJK7a+kVasfZfWrJslRY2tRCIwDxFLAJ/Y47jaKDnrM6daldH2/+q50yHCrmsfH6/kgtqlEhVKRD0fszJXAg4ZVXZ/aqy+iYDMm0P8azHWVA0Yhr11yjrr9FIoEZZf61G/BmZvvAV7c8yTeWyg8pnh+XYtqLxpGrrF+vm1f/U0uWrCUQAkohF4Ibhuq5GBvvUdapdp04e19mOn5XPjsixbYVjcSXSVYolUldt1PGPDNdR3MkqZQ8rOTGlnPHXmJNjJPIPHBkq+uLK+6omo/BCIJZ9sRnte/z6w7xyI0Mq5nMT/6mo1pJlq7Rk+Wotal6u6rpGrkEEMAWxCNygxvKj6jnTod4zp/T7yeMa7OvWWH5UhuFTLJlSIl015zfJTIffNScDMu6MKubkFXNyk68Rp3BDfXWhK6loxFXwp1TwpZX3p1XwpZT3pVXwp1XwJeUaszOa57quyqUxFUazGstl5diWwrG4qjMNaln1Ty1qXq7Gpctn/asnAdxYiEVgHnBdV8P9veo906Gu0/+nzl9/UnZoQJVyUT6/X7FESrFESpFY4prH458ZrqOIU/hDQOYVcosKOyWF3IvThfdBt3xV49KWX5YRlGUEVTYiKvtiKhlxlXxRlSbnYxPzMZV9MTmzFIN/drk4DEViSqar1LSsVY03LdfCxc3KNC7h18sApo1YBOYhyzTV192p82d/V8+ZDnX9/qvy2WGVigUZMhSOxRVLjgfkXHyTzJxyXQXdsoJuRX5Z8ru2fLLld63xSbZ8k6+2DLly5ZMrQ65hSDLkSuPvJ5f7xoNQQZlGSJZx8XW2RgGvhGPbKo7lVcznVCyMyrHtS+KwbtFS1dY3KRj6aze8AMAFxCIAWaapob5uDfSc1fmuTp3tOKmRgfMq5nNyHEeBUEjRWEKReELhaJxRqWvAcRyVi2MqjeVULORllssyfIYisYSS6So1Nt+ixqXLiEMAs45YBHAJ13WVGx5Uf88Z9fecU+/ZDvWd69RYIafSWF6u48jn9ysSTyoSiysaSygQJE5mg+u6skxT5VJhIg4LsioVGYYUisYUS6RU17hUjTctU01do2rqFytdU0fAA5gzxCKAabEtSyODfRoZ6NXwwHn195xVT2eH8qPDKo3lZVuWJCkQDCkciSkUiSociSoYjlyTO7Cvd5eNQnM8Cn3+oCLRmKLxpDKNTapbdJOq6xpUU9eoBbX1jBoCuKqIRQBXzHVd5UaGJgKyT7mRAQ30dmnwfJeK+VGVS0VVyiVJkiFDwXBEoUhEwVB4cvIHgtfdTTWzxbYsVcolVcpFmeWyKpWSrHJZ7sQNOP7A1CjMNCxRujqjdE1G6aqMEguqCW0A1xyxCGDWua6rsfyociODyg0PanRkSNmhfg32ntPIwHmVy0WZlYrMSlm2aerCgxZ9Pp8CExEZCATlDwTkDwTl90+8BgLXPCwdx5FlVmSbpiyrIsusyDLN8cmqyLEdGYYr15V8fr9C4ahC4bAisaSqautUVdeoZLpK8WRaiXQVUQjgukcsAriqxh8MXdBYflTF/KjGCqMay+cm4nJIo0P9Gh0eUKVckm2ZsixLtmXJti3ZljklFi/8+DIMQ4bhk+HzyefzTZ2/EGHuxHjexDYXf/S54/Ou5Di2HNuW49hyXWnyUK4rGRMvhk+BYFCBYEiBQFCBYFDRREqJ1AIlUuMRGI7FFU+kFE8tUCJdpXhygcLRa/9MSwC4EsQigOuSZZqqlIvjv8ouFWWWSyqXx+cr5dLkq+PYskxTZqU8McpXkVmpTM5blinpQlBOROVkXF587/f7FQxHFI7GFI7Exkc3gyEFgyEFQqHx+VBIwVBYkVhCkWhc4Whc4WiMm0sA3NCIRQAAAHjiIhkAAAB4IhYBAADgiVgEAACAJ2IRAAAAnohFAMCMnT59WoZh6NixY9f6VADMMmIRAOaBi48Ouvz01FNPzWj/TU1N6unp0a233npF2x8/flxbtmxRU1OTotGoWltbtWvXrhmdE4DZEbjWJwAAmHs9PT2T81988YV27typ9vb2yWXRaHRG+/f7/aqvr7/i7Y8ePapMJqPPPvtMTU1NOnz4sJ555hn5/X7t2LFjRucGYGYYWQSAeaC+vn5ySqfTMgxjyrLPP/9cN998s0KhkFasWKFPP/10yvaGYWj37t165JFHFI1G1dzcrH379k3++eV+Dd3W1qZHH31UqVRKyWRSGzZsUEdHx2XPb/v27Xr77be1ceNGtbS0aOvWrdq2bZv2798/J38fAKaPWASAee7AgQN6/vnn9dJLL+nEiRN69tlntW3bNn399ddT1nv99df1+OOP6/jx49q6dau2bNmikydPXnafXV1duvfeexWJRHTo0CEdPXpU27dvl2VZ0z6vbDar6urqGX02ADPHN7gAwDzz8ccf64UXXtDIyIgkaf369Vq9erXef//9yXU2b96sQqGggwcPShofWXzuuee0e/fuyXXuuusurVu3Tu+++65Onz6t5uZm/fjjj1q7dq1effVV7d27V+3t7QoGg3/5HL/77jtt3LhRBw8e1EMPPTSzDwxgRhhZBIB57uTJk1q/fv2UZevXr79k1PDuu+++5L3XyOKxY8e0YcOGKwrFtrY2PfbYY9q5cyehCFwHiEUAgAzDmPLedd1Llk1nuwuu9IaZn3/+WQ888ICefvppvfbaa1e0DwCzi1gEgHmutbVV33777ZRlhw8fVmtr65RlR44cueT9ypUrL7vPNWvW6JtvvpFpmtM+j7a2Nt1///168skn9eabb057OwBzi0fnAMA89/LLL2vz5s1at26dHnzwQX355Zfav3+/vvrqqynr7du3T7fffrvuuece7dmzR99//70+/PDDy+5zx44deuedd/TEE0/olVdeUTqd1pEjR3TnnXdqxYoVl6x/IRQffvhhvfjii+rt7ZU0/kieTCYz+x8awLQxsggA89ymTZu0a9cuvfXWW1q9erXee+89ffTRR7rvvvumrPfGG29o7969WrNmjT755BPt2bNHq1atuuw+a2pqdOjQIeXzeW3cuFG33XabPvjgA89rGPft26f+/n7t2bNHDQ0Nk9Mdd9wx2x8XwF/E3dAAgP/KMAwdOHBAmzZtutanAuAqY2QRAAAAnohFAAAAeOIGFwDAf8UVS8D8xcgiAAAAPBGLAAAA8EQsAgAAwBOxCAAAAE/EIgAAADwRiwAAAPBELAIAAMATsQgAAABPxCIAAAA8EYsAAADwRCwCAADAE7EIAAAAT8QiAAAAPBGLAAAA8EQsAgAAwBOxCAAAAE/EIgAAADwRiwAAAPBELAIAAMATsQgAAABPxCIAAAA8EYsAAADwRCwCAADAE7EIAAAAT8QiAAAAPBGLAAAA8EQsAgAAwBOxCAAAAE/EIgAAADwRiwAAAPBELAIAAMATsQgAAABPxCIAAAA8EYsAAADwRCwCAADAE7EIAAAAT8QiAAAAPBGLAAAA8EQsAgAAwBOxCAAAAE/EIgAAADwRiwAAAPBELAIAAMATsQgAAABP/w/HAGz++Q8aFgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "from sklearn.manifold import TSNE\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "# Apply t-SNE to reduce the dimensionality of the data\n",
    "tsne = TSNE(n_components=2, verbose=1, perplexity=40, n_iter=300)\n",
    "tsne_results = tsne.fit_transform(W)\n",
    "\n",
    "# Calculate the proportion of each topic\n",
    "topic_proportions = W.sum(axis=0) / W.sum()\n",
    "\n",
    "\n",
    "\n",
    "# Create a pie chart with the topic proportions\n",
    "fig, ax = plt.subplots(figsize=(8, 8))\n",
    "ax.pie(topic_proportions, labels=['Topic {}'.format(i+1) for i in range(len(topic_proportions))],\n",
    "       autopct='%1.1f%%', shadow=True, startangle=90)\n",
    "ax.set_title('Topic Proportions')\n",
    "\n",
    "plt.show()\n",
    "\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.17"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}
